Tables in Signals and Systems

Division of Signal Processing

Last revised October 1999

Contents

I Continuous-time Fourier series 2

I-A Properties of Fourier series . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
I-B Fourier series table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

II Continuous-time Fourier transform 4

II-A Properties of the Fourier transform . . . . . . . . . . . . . . . . . . . . . . . 4
II-B Fourier transform table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

III Discrete-time Fourier series 7

III-AProperties of discrete-time Fourier series . . . . . . . . . . . . . . . . . . . . 7
III-B Fourier series table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

IV Discrete-time Fourier transform 9

IV-A Properties of the discrete-time Fourier transform . . . . . . . . . . . . . . . 9
IV-B Discrete-time Fourier transform table . . . . . . . . . . . . . . . . . . . . . . 10

V Sampling and reconstruction 11

VI Z-transform 12
VI-A Properties of the Z-transform . . . . . . . . . . . . . . . . . . . . . . . . . . 12
VI-B Z-transform table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 2

Definitions

4 sin (πt) 4 2π
sinc (t) = Ωo =
πt T0
I. Continuous-time Fourier series
A. Properties of Fourier series
Periodic signal Fourier serie coefficient
∞ Z
X 4 1
x (t) = ak ejkΩo t ak = x (t) e−jkΩo t dt
To To
k=−∞

x(t) Periodic with ak
y(t) periodT0 bk

Ax (t) + By (t) Aak + Bbk

x (t − t0 ) ak e−jk(2π/T0 )t0

ejM (2π/T0 )t x (t) ak−M

x∗ (t) a∗−k
x (−t) a−k
x (αt) , α > 0 (Periodic with period T0 /α) ak
Z
x (τ ) y (t − τ ) dτ T0 ak bk
T0
∞
X
x (t) y (t) al bk−l
l=−∞
d 2π
x (t) jk ak
dt T0
Z t
1
x (τ ) dτ (Bounded and periodic only if a0 = 0) ak
−∞ jk (2π/T0 )
If x (t) is real valued then
ak = a∗−k


 < {ak } = < {a−k }

x (t) = {ak } = −= {a−k }
 |ak | = |a−k |


arg {ak } = − arg {a−k }
xe (t) = E{x(t)} <{ak }
xo (t) = O{x(t)} j={ak }

ak ejkΩ0 t + a−k e−jkΩ0 t = 2<{ak } cos(kΩ0 t) − 2={ak } sin(kΩ0 t)

Parsevals relation for periodic signals
Z ∞
1 2
X
|x(t)| dt = |ak |2
T0 T0
k=−∞
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 3

B. Fourier series table

x(t) ak or the Fourier series expansion
P∞ 1
a) n=−∞ δ(t − nT ) ak = , all k
T
b) 1 (a0 = 1, ak = 0 otherwise), ∀ T0 > 0
c) ejΩo t a1 = 1, ak = 0 otherwise
1
d) cos Ωo t a1 = a−1 = , ak = 0 otherwise
2
1
e) sin Ωo t a1 = −a−1 = , ak = 0 otherwise
2j
1, | t |< T1
(
To Ωo T1 sin kΩo T1
f) 0, T1 <| t |≤ ak = sinc kΩπo T1 =
2 π kπ
period T0
1, 0<t<π
n
4 sin t sin 3t sin 5t


g) −1, −π < t < 0 π 1 + 3 + 5 + ···

t, 0<t<π
n
π 4 cos t cos 3t cos 5t


h) |t| = −t, −π < t < 0 2 − π 12
+ 32
+ 52
+ ···

sin t sin 2t sin 3t

i) t, −π <t<π 2 1 − 2 + 3 − ···
sin t sin 2t sin 3t


j) t, 0 < t < 2π π−2 1 + 2 + 3 + ···
2 4 cos 2t cos 4t cos 6t


k) | sin t|, −π <t<π π − π 1·3 + 3·5 + 5·7 + ···
(
0, 0<t<π−a
a 2 sin a cos t sin 2a cos 2t sin 3a cos 3t


l) 1, π − a < t < π + a π − π 1 − 2 + 3 − ···
0, π + a < t < 2π
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 4

II. Continuous-time Fourier transform

A. Properties of the Fourier transform

Non-periodic signal Fourier transform

Z ∞ Z ∞
1 4
x(t) = X(jΩ)ejΩt dΩ X(jΩ) = x(t)e−jΩt dt
2π −∞ −∞

Alternativly with frequency f instead of angular

Z ∞ frequency Ω.
Z ∞
4
x(t) = Xf (f )e j2πf t df Xf (f ) = x(t)e−j2πf t dt = X(ω){Ω=2πf }
−∞ −∞

x(t) X(jΩ)
y(t) Y (jΩ)
ax(t) + by(t) aX(jΩ) + bY (jΩ)
x(t − t0 ) e−jΩt0 X(jΩ)
ejΩ0 t x(t) X(j(Ω − Ω0 ))
x∗ (t) X ∗ (j(−Ω))
x(−t) X(j(−Ω))
 
1 Ω
x(at) X
|a| a
x(t) ∗ y(t) X(jΩ)Y (jΩ)
1
x(t)y(t) X(jΩ) ∗ Y (jΩ)
2π
d
x(t) jΩX(jΩ)
dt
Z t
1
x(t)dt X(jΩ) + πX(0)δ(Ω)
−∞ jΩ
d
tx(t) j X(jΩ)
dΩ
If x(t) is real valued
 then
 X(jΩ) = X ∗ (j(−Ω))
 <{X(jΩ)} = <{X(j(−Ω))}

x(t) ={X(jΩ)} = −={X(j(−Ω))}
 |X(jΩ)| = |X(j(−Ω))|


arg{X(jΩ)} = − arg{X(j(−Ω))}
xe (t) = E{x(t)} <{X(jΩ)}
xo (t) = O{x(t)} j={X(jΩ)}
Duality
Z ∞ F
g(t) ←→ f (jΩ)
f (u) = g(v)e−juv dv, F
−∞ f (t) ←→ 2πg(j(−Ω))
Parsevals
Z ∞ relation for non-periodic
Z ∞ signals
2 1 2
|x(t)| dt = |X(jΩ)| dΩ
−∞ 2π −∞
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 5

B. Fourier transform table

The table is valid for <{α} > 0 and <{β} > 0

x(t) X(jΩ) X(f )

T

T

sin ΩT /2 sin πf T
a) u t+ 2 −u t− 2 T T = T sinc(f T )
ΩT /2 πf T
sin W t W Wt W

W


b) = sinc u (Ω + W ) − u (Ω − W ) u f+ 2π −u f − 2π
πt π π
|t|
(
T T

sin ΩT /4
2
T
c) 1−2 , | t |< sinc2 (T f /2)
T 2 2 ΩT /4 2
0, otherwise
1 1
d) e−αt u(t)
jΩ + α j2πf + α
2α 2α
e) e−α|t|
Ω2 + α2 (2πf )2 + α2
1  −αt 1 1
− e−βt u(t)

f) e
β−α (jΩ + α)(jΩ + β) (j2πf + α)(j2πf + β)
1 1
g) te−αt u(t)
(jΩ + α)2 (j2πf + α)2
tn−1 −αt 1 1
h) e u(t)
(n − 1)! (jΩ + α)n (j2πf + α)n
√ √
2 π −(Ω/2α)2 π −(πf /α)2
i) e−(αt) e e
α α
Ωo Ωo
j) e−αt sin(Ωo t)u(t)
(jΩ + α)2 + Ω2o (j2πf + α)2 + Ω2o
−Ωo −Ωo
eαt sin(Ωo t)u(−t)
(α − jΩ)2 + Ω2o (α − j2πf )2 + Ω2o
α + jΩ α + j2πf
k) e−αt cos(Ωo t)u(t)
(jΩ + α)2 + Ω2o (j2πf + α)2 + Ω2o
α − jΩ α − j2πf
eαt cos(Ωo t)u(−t)
(α − jΩ)2 + Ω2o (α − j2πf )2 + Ω2o
"
T sin(Ω − Ωo ) T2 T

sin πT (f − fo )
2 (Ω − Ωo ) T2 2 πT (f − fo )
T T



l) (cos Ωo t) u t + 2 −u t− 2 # 
sin(Ω + Ωo ) T2 ) sin πT (f + fo )
+ +
(Ω + Ωo ) T2 ) πT (f + fo )
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 6

Generalized Fourier transform (power signals)

x(t) X(jΩ) X(f )
a) δ(t) 1 1

b) δ(t − t0 ) e−jΩt0 e−j2πf t0

∞ ∞ ∞
X 2π X 2π 1 X n
c) δ(t − nT ) δ(Ω − n) δ(f − )
n=−∞
T n=−∞ T T n=−∞ T

1 1 1
d) u(t) πδ(Ω) + δ(f ) +
jΩ 2 j2πf
t 2 1
e) sgn(t) =
|t| jΩ jπf
1
f) −jsgn(Ω) −jsign(f )
πt
g) K 2πKδ(Ω) Kδ(f )
1 j 0 1
h) tu(t) jπδ 0 (Ω) − δ (f ) − 2 2
Ω2 4π 4π f
 n
j
i) tn 2π(j)n δ (n) (Ω) δ (n) (f )
2π
1
j) cos Ωo t π[δ(Ω − Ωo ) + δ(Ω + Ωo )] [δ(f − fo ) + δ(f + fo )]
2
π 1
k) sin Ωo t [δ(Ω − Ωo ) − δ(Ω + Ωo )] [δ(f − fo ) − δ(f + fo )]
j j2
∞ ∞   ∞
X 2π X 2πn X
cn ej T nt n


l) 2π cn δ Ω − cn δ f − T
n=−∞ n=−∞
T n=−∞

m) ejΩo t 2πδ(Ω − Ωo ) δ(f − fo )

n) Periodic square wave
∞ ∞
1, | t |≤ T1
( X X
Ak (Ω)δ (Ω − kΩo ) Ak (f )δ (f − kfo )
To
0, T1 <| t |≤ k=−∞ k=−∞
2 2 sin kΩo T1 sin k2πfo T1
periodTo Ak (Ω) = Ak (f ) =
k kπ
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 7

III. Discrete-time Fourier series

A. Properties of discrete-time Fourier series
Periodic signal Fourier serie coefficient
X 4 1 X
x[n] = ak ejk(2π/N )n ak = x[n]e−jk(2π/N )n
N
k=<N > n=<N >

x[n] Periodic with ak Periodic with
o
y[n] period N bk period N

Ax[n] + By[n] Aak + Bbk

x[n − n0 ] ak e−jk(2π/N )n0

ejM (2π/N )n x[n] ak−M

x∗ [n] a∗−k
x[−n] a−k
n
x[n/m], If n is a multiple av m 1
x(m) [n] = 0, otherwise ak , period mN
m
X
x[r]y[n − r] N ak bk
r=<N >
X
x[n]y[n] al bk−l
l=<N >
1 − e−j2π/N k ak


x[n] − x[n − 1]
n
X Bounded and periodic 1
x[k] only if a = 0 ak
k=−∞
0 1 − e−jk2π/N

If x[n] is real valued then

ak = a∗−k


 <{ak } = <{a−k }

x[n] ={ak } = −={a−k ]}
 |ak | = |a−k |


arg{ak } = − arg{a−k }
xe [n] = E{x[n]} <{ak }
xo [n] = O{x[n]} j={ak }
Parsevals relation for periodic signals
1 X X
|x[n]|2 = |ak |2
N
n=<N > k=<N >
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 8

B. Fourier series table

x[n] ak
∞
X 1
δ(n − kN ) ak = ,for all k
N
k=−∞

1, k=0,±N,±2N,...
n
1 ak = 0, otherwise


2πm

 ωo =
N



(
1, k=m,m±N,m±2N,...
ejωo n



 ak =
 0, otherwise
ωo
= irrational : The signal is non-periodic
2π

2πm

 ωo =
N



( 1
 2 , k=±m,±m±N,±m±2N,...
cos ωo n
 ak =



0, otherwise
ωo
= irrational : The signal is non-periodic
2π

2πm

 ωo =
N




 1
  2j , k=m,m±N,m±2N,...
sin ωo n 

 ak =

  −1, k=−m,−m±N,−m±2N,...
2j

ωo
= irrational : The signalen is non-periodic
2π

2πk 1


sin N N 1 + 2
1, | n |≤ N1
(
,

 k6=0,±N,±2N,...
N sin πk

N

ak = N
0, N1 <| n |≤
2
 2N1 + 1 ,

period N


k=0,±N,±2N,...
N
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 9

IV. Discrete-time Fourier transform

A. Properties of the discrete-time Fourier transform
Non-periodic signal Fourier transform
Z ∞
1 4 X
x[n] = jω jωn
X(e )e dω jω
X(e ) = x[n]e−jωn
2π 2π n=−∞
 
x[n] X(ejω ) Periodic with
y[n] Y (ejω ) period 2π

ax[n] + by[n] aX(ejω ) + bY (ejω )

x[n − n0 ] e−jωn0 X(ejω )

ejω0 n x[n] X(ej(ω−ω0 ) )

x∗ [n] X ∗ (ej(−ω) )
x[−n] X(ej(−ω) )
x[n/m], n multiple of m
n
x(m) [n] = 0, n not multiple av m X(ej(mω) )

x[n] ∗ y[n] X(ejω )Y (ejω )

Z
1
x[n]y[n] X(ejθ )Y (ej(ω−θ) )dθ
2π 2π
1 − ejω X(ejω )


x[n] − x[n − 1]
n ∞
X 1 jω ) + πX(0)
X
x[k] X(e δ(ω − 2πk)
1 − ej ω
k=−∞ k=−∞

d
nx[n] j X(ejω )
dω
If x[n] is real valued
 then

 X(ejω ) = X ∗ (ej(−ω) )
 <{X(ejω )} = <{X(ej(−ω) )}


x[n] ={X(ejω )} = −={X(ej(−ω) )}
X(ejω ) = X(ej(−ω) )




arg{X(ejω )} = − arg{X(ej(−ω) )}


xe [n] = E{x[n]} <{X(ejω )}

xo [n] = O{x[n]} j={X(ejω )}
Parsevals relation for non-periodic signals
∞ Z
X
2 1
|x[n]| = |X(ejω )|2 dω
n=−∞
2π 2π
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 10

B. Discrete-time Fourier transform table

x[n] X(ejω )
δ[n] 1
δ[n − n0 ] e−jωn0
∞ ∞  
X 2π X 2πk
δ(n − kN ) δ ω−
N N
k=−∞ k=−∞
∞
X
1 2π δ (ω − 2πk)
k=−∞
∞
X
ejωo n 2π δ (ω − ωo − 2πk)
k=−∞
∞
X
cos ωo n π [δ(ω − ωo − 2πk) + δ(ω + ωo − 2πk)]
k=−∞
∞
π X
sin ωo n [δ(ω − ωo − 2πk) − δ(ω + ωo − 2πk)]
j
k=−∞
∞
1 X
u[n] +π δ(ω − 2πk)
1 − e−jω
k=−∞

1
an u(n), |a| < 1
1 − ae−jω
1
(n + 1)an u[n], |a| < 1
(1 − ae−jω )2
(n + m − 1)! n 1
a u[n], |a| < 1
n!(m − 1)! (1 − ae−jω )m
1 1
a|n| , |a| < 1
1 − a2 1+ a2 − 2acosω
1, |n| ≤ N1
(
∞  
N X 2πk
0, N1 < |n| ≤ 2π ak δ ω
2 N
period N k=−∞

1


1, |n| ≤ N1 sin ω N1 + 2
0, |n| > N1 sin ω2

sin W n W Wn
(
1, |ω| ≤ W
= sinc 0, W < |ω| ≤ π
πn π π
0<W <π period 2π
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 11

V. Sampling and reconstruction

The sampling theorem:
Let x(t) with transform Xc (jΩ) be a bandlimited signal such that Xc (jΩ) = 0, |Ω| > ΩM .
Then x(t) is uniquely described by the samples x(nT ), n = 0, ±1 ± 2 . . . if

Ωs > 2ΩM

where
2π
Ωs =
= 2πfs
T
Given x(nT ), if the sampling theorem is satisfied, it is possible with an ideal reconstruction
filter to exactly reconstruct x(t).

Discrete-time processing of continuous-time signals

Sampling:
∞
X ∞
X
xd (t) = x(nT )δ(t − nT ) ←→ Xd (jΩ) = x(nT )e−jΩnT
n=−∞ n=−∞

”Normalization in time” gives

∞
X ∞
X
x[n] = x(nT ) ←→ X(ejω ) = x[n]e−jωn = x(nT )e−jΩnT
n=−∞ n=−∞

where
Ω f
ΩT = ω = or f T = q =
fs fs
Poissons summation formula:
∞
1 X
Xd (jΩ) = Xc (j(Ω − kΩs ))
T
k=−∞

∞
jω 1 X Ω 2πk
X(e ) = Xc (j( − ))
T T T
k=−∞

If the sampling theorem is satisfied then

1 π π
Xd (jΩ) = X(jΩ), − < Ω <
T T T
or
1 1 1
Xd (f ) = X(f ), − <f <
T 2T 2T
Ideal reconstruction:
∞
X
xr (t) = x(nT )h(t − nT )
n=−∞

where
Ωc Ωc t | ω |≤ Ωc
←→ H(jΩ) = T,
n
h(t) = T sinc 0, otherwise
π π
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 12

VI. Z-transform
A. Properties of the Z-transform
signal Z-transform ROC
∞
4 X
x[n] X(z) = x[n]z −n Rx
n=−∞

ax[n] + by[n] aX(z) + bY (z) Contains Rx ∩ Ry

x[n − n0 ] z −n0 X(z) Rx , except possible addition
or deletion of the origin or ∞

z0n x[n] X(z/z0 ) |z0 |Rx

x∗ [n] X ∗ (z ∗ ) Rx
x∗ [−n] X ∗ (1/z ∗ ) 1/Rx
x[n] ∗ y[n] X(z)Y (z) Contains Rx ∩ Ry
d
nx[n] −z X(z) Rx , except possible addition
dz
or deletion of the origin or ∞

<{x[n]} 1
2 [X(z) + X ∗ (z ∗ )] Contains Rx

={x[n]} 1
2j [X(z) − X ∗ (z ∗ )] Contains Rx

Initial value theorem

x[n] = 0, n < 0 limz→∞ X(z) = x[0]
TABLES IN SIGNALS AND SYSTEMS, OCT. 1999 13

B. Z-transform table
x[n] X(z) ROC
δ[n] 1 All z
δ[n − n0 ] z −n0 All z, except 0(n0 > 0)
or ∞(n0 < 0)
1
u[n] |z| > 1
1 − z −1
1
−u[−n − 1] |z| < 1
1 − z −1
1
an u[n] |z| > a
1 − az −1
1
−an u[−n − 1] |z| < a
1 − az −1
az −1
nan u[n] |z| > a
(1 − az −1 )2
az −1
−nan u[−n − 1] |z| < a
(1 − az −1 )2
1 − [cos ω0 ]z −1
[cos ω0 n]u[n] |z| > 1
1 − [2 cos ω0 ]z −1 + z −2
1 − [sin ω0 ]z −1
[sin ω0 n]u[n] |z| > 1
1 − [2 cos ω0 ]z −1 + z −2
1 − [r cos ω0 ]z −1
[rn cos ω0 n]u[n] |z| > r
1 − [2r cos ω0 ]z −1 + r2 z −2
1 − [r sin ω0 ]z −1
[rn sin ω0 n]u[n] |z| > r
1 − [2r cos ω0 ]z −1 + r2 z −2
n
an , 0 ≤ n ≤ N − 1 1 − aN z −N
0, otherwise |z| > 0
1 − az −1