DISCRETE-TIME SIGNALS AND SYSTEMS

1 Discrete-Time Signals

 Recall we are interested in discrete-time continuous-amplitude signals

Continuous amplitude Discrete amplitude

Analog signal

Continuous-time

Digital signal

Discrete-time

Fig. 1 Classification of signals

 Strictly, a discrete-time signal is simply a sequence of numbers, represented

mathematically as follows
x = { x[n ]} (1)

where n Î  and x[n] is nth number of the sequence

 However, in digital signal processing, x often results from sampling an analog

signal xc (t ) periodically, i.e.

x [n ] = xc (nT ), nÎ (2)

where T is sampling period

Curtin University, Australia 2-1

YH Leung (2005, 2006, 2007, 2012, 2018)
 Sample index n thus has connotation of time. Suppose xc (t ) was sampled from
t = n1T to t = n2T . We then write
n2
x = { x [n ]} (3)
n = n1

By convention, it will be assumed x[n] = 0 for n < n1 and for n > n2 . x is then
said to be a finite length sequence

 For very long sequences, possibly infinite in length, we use also following
notation whose meaning is self-evident

x = {, x0 , x1, x2 , x3 , }

n=0 (4)

 Remarks
(i) As stated, strictly, x[n] means “nth number of the sequence x”
However, by corruption of use, it can also refer to the entire sequence itself

(ii) x[n] is defined only for integer values of n

(iii) x[n] is assumed to be always finite in value, i.e.

x [n ] < ¥ , "n Î  (5)

Curtin University, Australia 2-2

YH Leung (2005, 2006, 2007, 2012, 2018)
2 Some Basic Sequences

 Unit sample sequence (aka discrete-time impulse, or simply, impulse)

ìï0, n ¹ 0
d[n ] = ïí (6)
ïïî1, n = 0

d [n]

… …
-3 -2 -1 0 1 2 3 4 n

 Unit step sequence

ïì0, n < 0
u[n ] = ïí (7)
ïïî1, n ³ 0

u[n]

1
…
…
-3 -2 -1 0 1 2 3 4 n

 Exponential sequence

x [n ] = a n , a Î  (8)

x[n] x[n]

… 1 0<a<1 … 1 -1 < a < 0

… …
-3 -2 -1 0 1 2 3 4 n -3 -2 -1 0 1 2 3 4 n

Curtin University, Australia 2-3

YH Leung (2005, 2006, 2007, 2012, 2018)
 Alternating sequence (exponential sequence with a = -1 )

x[n ] = (-1)n = {, + 1, - 1, + 1, - 1, + 1, } (9)


n =0

x[n]

… …
-3 -2 -1 0 1 2 3 4 n

 Periodic sequence

The sequence x[n] is said to be periodic with period N Î  if

x [n + N ] = x [n ], "n Î  (10)

Fundamental period of x[n] is smallest positive N that satisfies Eq. (10)1

 Sinusoidal sequence

x [n ] = A cos(qo n + f ), A, qo , f Î  (11)

A is amplitude, qo is frequency, f is phase

The sinusoidal sequence x[n] is periodic with period N Î  if and only if

qo N = 2pk , for some k Î (12)

 Complex exponential sequence

x [n ] = Aa n = A e j f a n e j qo n
x [n ] = Aa n = A a n {cos(qo n + f ) + j sin(qo n + f )} (13)

where A = A e jf , a = a e j qo , A, a Î  , f, qo Î 

1
Thus alternating sequence has fundamental period N = 2

Curtin University, Australia 2-4

YH Leung (2005, 2006, 2007, 2012, 2018)
3 Basic Sequence Operations

 Suppose x = { x [n ]} and y = {y [n ]} are two sequences

Scaling: a ⋅ x = {a ⋅ x [n ]}, a Î  (14)

Addition2: x + y = { x [n ] + y [n ]} (15)

Multiplication2: x ⋅ y = { x [n ]⋅ y [n ]} (16)

¥ ¥
Convolution: x*y = å x [k ] y [n - k ] = å x [n - k ] y [k ] (17)
k =-¥ k =-¥

Example 1

(a) x = {1, 2, 3, 4}3n=0

y = {5, 6, 7}3n=1

n 0 1 2 3
x 1 2 3 4
y 5 6 7
2x 2 4 6 8
x+y 1 7 9 11
x⋅y 0 10 18 28

Therefore 2 x = {2, 4, 6, 8}3n=0

x + y = {1, 7, 9, 11}3n=0

x ⋅ y = {10, 18, 28}3n=1

2
Thus, sum and product of two sequences are given, respectively, by their sample-by-sample sum
and product

Curtin University, Australia 2-5

YH Leung (2005, 2006, 2007, 2012, 2018)
 ì 0,
ï n<0
(b) x[n ] = ï
í n can be written more compactly as x [n ] = Aa n u [n ]
ï
ï
î Aa , n ³ 0

Aαn
A

0 n

u[n]

0 n
point-by-point
n multiply
Aα u[n]

0 n

(c) x[n ] = {1, 4, 0, - 1}2n=-1 can be written mathematically as x [n ] = d[n + 1] + 4d[n ] - d[n - 2]

d[ n ]
1

0 n

d [ n - 2]
1
n-2 = 0  n=2

0 2 n

d [ n + 1]
1
n +1= 0  n = -1

-1 0 n

 Example 1(c) yields the general decomposition

¥
x [n ] = å x [k ] d [n - k ] (18)
k =-¥

Curtin University, Australia 2-6

YH Leung (2005, 2006, 2007, 2012, 2018)
Example 1 (cont.)
2 2
(d) Suppose x[n ] = {1, 4, 0, - 1}n=-1 and y [n ] = {1, 0, - 1}n=0 . Define w [n ] = x [n ] * y [n ]

n = -1 w [-1] = x [-1] y [0] = 1⋅ 1 = 1

n=0 w [0] = x [-1] y [1] + x [0] y [0] = 1⋅ 0 + 4 ⋅ 1 = 4
n =1 w [1] = x [-1] y [2] + x [0] y [1] + x [1] y [0] = 1⋅ (-1) + 4 ⋅ 0 + 0 ⋅ 1 = - 1
n=2 w [2] = x [0] y [2] + x [1] y [1] + x [2] y [0] = 4 ⋅ (-1) + 0 ⋅ 0 + (-1) ⋅ 1 = - 5
n=3 w [3] = x [1] y [2] + x [2] y [1] = 0 ⋅ (-1) + (-1) ⋅ 0 = 0
n=4 w [4] = x [2] y [2] = (-1) ⋅ (-1) = 1

Or more conveniently, by following Table method

x [n ]
1 4 0 -1
y [n ]
1 1 4 0 -1

0 0 0 0 0

-1 -1 -4 0 1

Adding along indicated diagonals yields

w [n ] = {1, 4, - 1, - 5, 0, 1}n4=-1

 It can be shown that (see Tutorial)

M2
x = { x[n ]}
n =M1 M 2 + N2
 x * y = {w [n ]} (19)
N2 n =M1+N1
y = {y [n ]}
n =N1

Curtin University, Australia 2-7

YH Leung (2005, 2006, 2007, 2012, 2018)
4 Discrete-Time Fourier Transform (DTFT)

 Discrete-time Fourier transform of x[n] is defined by

¥
DTFT { x[n ]} = X (e ) = jq
å x [n ] e- jnq , q Î  (20)
n =-¥

 X (e j q ) is periodic in q with period 2p

Proof
X (e j ( q +2mp ) ) = X (e j q e j 2mp ) = X (e j q ), "m Î  (21)

Therefore, we consider only q Î [-p, p ]

 We call q digital frequency. Its unit is radians/sample

 Inverse DTFT is given by

DTFT-1 { X (e j q )} = x [n ] =
1
X (e j q ) e+ jnq d q
2p ò2p
(22)

where integral is evaluated over any 2p interval of q

 For Example 2, following result on geometric series is useful

ìï1 - aN
N -1 ïï , a ¹1
å a = ïíï 1 - a
n
(23)
n =0 ïï
ïî N , a =1

Curtin University, Australia 2-8

YH Leung (2005, 2006, 2007, 2012, 2018)
Example 2
(a) x [n ] = d [n ]
¥
X (e ) =
jq
å d[n ]e
- jnq
=1
n =-¥

(b) x [n ] = a n u[n ], a <1

¥ ¥ n 1
jq
X (e ) = å n
a u[ n ]e
- jnq
= å (ae- j q ) =
- jq
n =-¥ n =0 1 - ae

1
1- a

1
1+ a
a = 0.75

tan-1( a 1- a 2 )

1
1+ a

1
1- a

a = -0.75
tan-1( a 1- a 2 )

Curtin University, Australia 2-9

YH Leung (2005, 2006, 2007, 2012, 2018)
(c) x [n ] = a n , a <1
¥ ¥ -1 ¥ n ¥
jq m
jq
X (e ) = å a e
n - jnq
= å an e- jnq + å a
-n - jnq
e = å (ae- j q ) + å ae ( )
n =-¥ n =0 n =-¥ n =0 m =1
jq
1 ae
= +
- jq jq
1 - ae 1 - ae
1- a 2
=
1 - 2a cos q + a 2

1+ a
1- a

1- a
1+ a

a = 0.75

(d) x [n ] = - a n u[-n - 1], a <1

¥ -1 n ¥
-1 j q n
jq
X (e ) = - å n
a u [-n - 1]e
- jnq
= - å (ae- j q ) = -å a ( e )
n =-¥ n =-¥ n =1
é¥ -1 j q n
ù
ê ( )
= - êå a e -1ú
ú
ë n =0 û

Last summation converges only if a-1e j q < 1 or a-1 < 1 which contradicts the given condition
a < 1 . Therefore, given sequence does not have a DTFT

?
a = 0.75

Curtin University, Australia 2-10

YH Leung (2005, 2006, 2007, 2012, 2018)
 Table of DTFT pairs

Sequence x [ n ] Transform X (e j q )

1. d [n ] 1
¥
1
2. u [n ]
1 - e- j q
+ å pd(q - 2pk )
k =-¥

3. d[n - m ] e- j q m
1
4. a n u [n ] , a <1
1 - ae- j q

1
5. ( n + 1)a n u[n ] , a <1
(1 - ae- j q )2

(n + r - 1)! n 1
6. a u[n ] , a <1
n !(r - 1)! (1 - ae- j q )r

ìï1,
ïí
n £ N1 sin(q(N1 + 21 ))
7.
ïï0 , n > N1 sin(q 2)
î

sin qc n q æq n ö ïìï1, q £ qc
8. = c sinc çç c ÷÷÷ í
pn p çè p ø ïï0 , qc < q £ p
î
¥
9. 1 2p å d(q - 2pk )
k =-¥

¥
10. e j qo n 2p å d(q - qo - 2pk )
k =-¥

¥
11. cos qo n p å {d (q - qo - 2pk ) + d(q + qo - 2pk )}
k =-¥

p ¥
12. sin qo n å {d(q - qo - 2pk ) - d(q + qo - 2pk )}
j k =-¥

¥
13. Periodic square wave
2p å ak d (q - 2Npk )
ïìï1, n £ N1 k =-¥
í
ïï0 , N < n £ N 2
î 1 sin éë 2Npk (N1 + 21 )ùû
ak =
x [n + N ] = x [n ] N sin éë 22pNk ùû

14. Impulse train

2p ¥
¥
å d(n - kN )
å d (q - 2Npk )
N k =-¥
k =-¥

Curtin University, Australia 2-11

YH Leung (2005, 2006, 2007, 2012, 2018)
 Properties of DTFT

Let X (e j q ) and Y (e j q ) be the DTFTs of x [n] and y [n] respectively:

 X (e j q )
DTFT
x [n ] ¬¾¾¾

 Y (e j q )
DTFT
y [n ] ¬¾¾¾

 aX (e j q ) + bY (e j q )
DTFT
1. Linearity ax [n ] + by [n ] ¬¾¾¾

 X (e j q ) e- j qno
DTFT
2. Time shifting x [n - no ] ¬¾¾¾

x [n ] e j qo n ¬¾ ¾¾
 X (e j ( q-qo ) )
DTFT
3. Frequency shifting

x * [n ] ¬¾ ¾¾
 X * (e- j q )
DTFT
4. Conjugation

 X (e- j q )
DTFT
5. Time reversal x [-n ] ¬¾ ¾¾

ìï x [n k ], n = multiple of k
x( k ) [n ] = ïí  X (e jk q )
DTFT
6. Time expansion ¬¾¾¾
ïïî 0, n ¹ multiple of k

 X (e j q )Y (e j q )
DTFT
7. Convolution x [n ] * y [n ] ¬¾ ¾¾

1
X (e j x )Y (e j (q-x ) ) d x
2p ò2p
DTFT
8. Multiplication x[n ] y [n ] ¬¾¾¾


 (1- e- j q ) X (e j q )
DTFT
9. Time differencing x [n ] - x [n - 1] ¬¾¾¾

n ¥
1
10. Accumulation å DTFT
x [k ] ¬¾¾¾

1- e- j q
X (e j q ) + p X (e j 0 ) å d(q - 2pk )
k =-¥ k =-¥

d
X (e j q )
DTFT
11. Frequency differentiation n x [n ] ¬¾ ¾¾
 j
dq

12. x [ n ] real X ( e- j q ) = X * ( e j q )

13. x [ n ] real and even X (e j q ) real and even

14. x [ n ] real and odd X (e j q ) purely imag and odd

¥
1 2
15. Parseval’s relation å x[n ] 2 =
2p ò 2 p
X (e j q ) d q
n =-¥

¥
1
å x [ n ] y * [n ] =
2p ò2p
X (e j q )Y * (e j q ) d q
n =-¥

Curtin University, Australia 2-12

YH Leung (2005, 2006, 2007, 2012, 2018)
Example 3

x[n]

-4 -3 -2 -1 0 1 2 3 4 n

DTFT of above sequence can be found as follows

Method 1 Decompose x[n] as follows

-4 -3 -2 -1 0 1 2 3 4 n -4 -3 -2 -1 0 1 2 3 4 n

From Entry 7 of DTFT table and linearity property

sin q(4 + 21 ) sin q(1 + 21 )

X1(e j q ) = + 2⋅
sin (q 2) sin (q 2)

Method 2 Decompose x [n] as follows

1 1

-4 -3 -2 0 n -1 0 1 n 0 2 3 4 n

From Entry 7 of DTFT table, and time-shifting and linearity properties

sin q(1+ 21 ) sin q(1 + 21 ) sin q(1 + 21 )

X 2 (e j q ) = e + j 3q + 3 ⋅ + e- j 3q
sin (q 2) sin (q 2) sin (q 2)

sin q(1 + 21 ) sin q(1 + 21 )

= 2⋅ cos3q + 3 ⋅
sin (q 2) sin (q 2)

It can be shown (exercise) X1(e j q ) = X 2 (e j q )

Curtin University, Australia 2-13

YH Leung (2005, 2006, 2007, 2012, 2018)
5 z-Transform

 z-transform of x[n] is defined by the Laurent series

 { x [n ]} = X ( z )
=  + x[-2] z 2 + x[-1] z + x[0] + x [1] z-1 + x [2] z-2 +  (24)
¥
= å x[n ] z-n , zÎ
n =-¥

 Recall from Eq. (20)

¥
X (e j q ) = å x [n ] e- jnq , q Î 
n =-¥

Therefore, z-transform generalises DTFT:3

X (e j q ) is simply X(z) evaluated around the unit circle z = e j q

 Region in z-plane where X(z) is finite is its region of convergence (ROC)

 Clearly, if X (e j q ) exists, then ROC of X(z) must include unit circle

 ROCs consist of rings in z-plane centered about the origin

 Inverse z-transform is given by

1
 -1 { X ( z )} = x[n ] = òC X (z ) z
n -1
dz (25)
2p j

where C is a closed contour in ROC of X(z)

3
In the same way Laplace transform generalises the Fourier transform

Curtin University, Australia 2-14

YH Leung (2005, 2006, 2007, 2012, 2018)
Example 4

In the following, we re-work Example 2 for z-transforms

(a) x [n ] = d [n ]
¥
X (z) = å d[n ] z-n = 1 and converges everywhere
n =-¥

(b) x [ n ] = a n u [n ]

Im
¥ ¥
X (z) = å a n u[n ] z-n = å an z-n
n=-¥ n=0
unit circle
¥
= å (az -1 n
)
n=0
1
=
1 - az-1
a 1 Re

provided az-1 < 1 z-plane

i.e., z > a

(c) x [n ] = a n

¥ ¥ -1
X (z) = å a n z-n = å an z-n + å a-n z-n
n=-¥ n=0 n=-¥ Im
¥ ¥
= å (az-1)n + å (az )n
n=0 n=1
1 é 1 ù
= +ê - 1ú
1- az-1 êë 1- az úû
(1- az ) + (1- az-1 ) - (1- az-1 )(1- az ) 1/a
= -1
(1- az )(1- az ) a 1
Re
2
1- a
=
(1 + a ) - a( z-1 + z )
2 z-plane

provided az-1 < 1 and az < 1

i.e., a < z < 1

a

Curtin University, Australia 2-15

YH Leung (2005, 2006, 2007, 2012, 2018)
(d) x [n ] = - a n u[-n - 1]

Im
-1 ¥
X (z) = å -a n z-n = - å (a-1z )n
n =-¥ n =1
é 1 ù
= -ê - 1ú
êë 1- a-1z úû
1
=
1- az-1
a 1
Re

provided a-1z < 1 z-plane

i.e. z < a

 Comparing Examples 4(b) and 4(d), we see that

X(z) is not unique – different sequences can have the same X(z)

X(z)s distinguishable only through their ROCs

 Beware:
ROC of X(z), denoted by R X , is plotted in z-plane, not z-1-plane although
by convention, X(z) is given in terms of z-1

Curtin University, Australia 2-16

YH Leung (2005, 2006, 2007, 2012, 2018)
 Sequence ROC

x[n]

All z
except z = 0
1.
n

x[n]

All z
except z = 
2.
n

x[n]
All z
except z = 0 and 
3.
n

x[n]

R– < |z|
4.
n

x[n]

5. |z| < R+

x[n]

6. R– < |z| < R+

Fig. 2 Regions of convergence of z-transforms

Curtin University, Australia 2-17

YH Leung (2005, 2006, 2007, 2012, 2018)
 Table of z-transform pairs

Sequence Transform ROC

1. d [n ] 1 All z

1
2. u[ n ] z >1
1 - z-1

1
3. -u[-n - 1] z <1
1 - z-1

All z except 0 (if m > 0 )

4. d [n - m] z-m
or ¥ (if m < 0 )

1
5. a n u[n ] z > a
1 - az-1

1
6. -a n u[-n - 1] z < a
1 - az-1

az-1
7. n a n u[ n ] z > a
(1 - az-1 )2

az-1
8. -n an u[-n - 1] z < a
(1 - az-1 )2

1 - [cos qo ] z-1
9. [cos qo n ]u[n ] z >1
1 - [2cos qo ] z-1 + z-2

[sin qo ] z-1
10. [ sin qo n ]u[n ] z > 1
1 - [2cos qo ] z-1 + z-2

1 - [r cos qo ] z-1
11. éê r n cos qo n ùú u[n ] z > r
ë û 1 - [2r cos qo ] z-1 + r 2 z-2

[r sin qo ] z-1
12. éê r n sin qo n ùú u [n ] z > r
ë û 1 - [2r cos qo ] z-1 + r 2 z-2

ì
ïa n , 0 £ n £ N - 1 1 - aN z-N
13. ïí z > 0
ï
ï 1 - az-1
î 0, otherwise

Curtin University, Australia 2-18

YH Leung (2005, 2006, 2007, 2012, 2018)
 Properties of z-transform

Let X(z) with ROC R X and Y(z ) with ROC RY be z-transforms of x[n] and y[n] respectively:


x [n ] ¬¾¾
 X ( z ), R X

y [n ] ¬¾¾
 Y ( z ), RY


1. Linearity ax [n ] + by [n ] ¬¾¾
 aX ( z ) + bY ( z ), contains R X Ç RY

 X ( z ) z-no , R X except for possible addition or deletion


2. Time shifting x [n - no ] ¬¾¾
of origin or ¥

3. Frequency shifting x [n ] zon ¬¾¾

 X

( zz ),
o
zo R X

x * [n ] ¬¾¾
 X * ( z* ), R X

4. Conjugation

 X ( z-1 ), 1 R X

5. Time reversal x [-n ] ¬¾¾

ì
ï x [r ], n = rk
x( k ) [n ] = ï

6. Time expansion í ¬¾¾ X ( z k ), R1X k
ï
ï
î 0, n ¹ rk


7. Convolution x [n ] * y [n ] ¬¾¾
 X ( z )Y ( z ), contains R X Ç RY

òC X (x )Y ( zx ) x1 d x , contains R X Ç RY , C in ROC

 1
8. Multiplication x[n ] y [n] ¬¾¾

2p j

 (1- z-1 ) X ( z ), contains R X Ç { z > 0}


9. Time differencing x [n ] - x [n - 1] ¬¾¾

n
1
10. Accumulation å x [k ] ¬¾¾


1 - z -1
X ( z ), contains R X Ç { z > 1}
k =-¥

 d
11. Differentiation in z n x [n ] ¬¾¾
 -z X ( z ), R X
dz

12. x [ n ] real X ( z ) = X * ( z* )

¥
13. Parseval’s relation å x [n ] 2 =
1
2p j òC X (z ) X
*
( z1 ) z1 dz , C in R X
*
n=-¥

¥
å x [ n ] y * [n ] =
1
2p j òC X ( z )Y
*
( z1 ) z1 dz , C in R X Ç RY
*
n=-¥

Curtin University, Australia 2-19

YH Leung (2005, 2006, 2007, 2012, 2018)
 All z-transform properties can be proved from the definition Eq. (24), but proof
can be tedious
See Appendix I for proof of time-shifting property

 Meaning of zo R X , 1 R X and R1X k

 1
Consider x [n ] = a n u [n ] ¬¾¾
 X (z) = , RX = { z > a }
1 - a z -1

(a) Frequency Shifting Property. It follows from the definition Eq. (24) that
¥
1 1
x [n ]zon = (azo )n u[n ] ¬¾¾


å (azo )n z-n = 1 - az -1
= -1
n=0 oz 1- a ( zz )
o

provided azo z-1 < 1

i.e. zo R X = { z > azo } = { z > zo ⋅ a }

(b) Time Reversal Property. Likewise

0 0 ¥
1
x [-n ] = a-n u[-n ] ¬¾¾


å a-n z-n = å am zm = å (az )m = 1 - az
n=-¥ m=¥ m=0

provided az < 1
i.e. 1 RX = { z < 1 a }

(c) Time Expansion Property. Finally

ìï x [r ], n = rk
x( k ) [n ] = ïí x [0]z 0 + 0 ⋅ z-1 +  + 0 ⋅ z-( k -1)

¬¾¾

ïïî 0, n ¹ rk
+ x [1]z-k + 0 ⋅ z-( k +1) +  + 0 ⋅ z-(2k -1)

+ x [2]z-2k + 0 ⋅ z-(2k +1) +  + 0 ⋅ z-(3k -1) + 

¥
= å x[r ](zk )-r
r =0
¥
= å ar (zk )-r
r =0
1
= provided az-k < 1
1 - az-k

i.e. R1X k = {z > a1k }

Curtin University, Australia 2-20

YH Leung (2005, 2006, 2007, 2012, 2018)
 Linearity property states

ax [n ] + by [n ] ¬¾¾
 aX ( z ) + bY ( z ), contains R X Ç RY

Significance of word “contains” explained in Appendix II – due primarily to

pole-zero cancellation or sequence cancellation

 Significance of word “contains” in convolution, multiplication, time-differencing

and accumulation properties can be explained similarly

Example 5
n
(a) x[n] = 3n u[-n - 1] + ( 21 ) u[n ]

-1
Entry 6: 3n u[-n - 1] « , z < 3
1 - 3 z-1

n 1
Entry 5: ( 21 ) u [n ] «
1 z -1
, z > 1
2
1- 2

Linearity Property:
-1 1
X (z) = -1
+ , 1 < z <3
1 - 3z 1- 1 z-1 2
2

n
(b) x[n ] = (- 21 ) u[n - 1]

n n-1
Observe: x[n ] = (- 21 ) u[n - 1] = (- 21 )(- 21 ) u[n - 1]

n 1 1
Entry 5: (- 21 ) u [n ] «
1 z-1
= , z > - 21
1+ 2
1 - (- 21 )z-1

Time-Shifting Property:
n-1 1
(- 21 ) u[n - 1] «
1 z-1
⋅ z-1, z > 1
2
1+ 2

Linearity Property:

z-1
X ( z ) = (- 21 ) ⋅ , z > 1
1+ 1 z-1 2
2

Curtin University, Australia 2-21

YH Leung (2005, 2006, 2007, 2012, 2018)
Example 6
1 1
X ( z ) = X 1( z ) + X 2 ( z ) = -1
+ , 0.5 < z < 2
1 - 0.5z 1 - 2 z-1

Suppose X1( z ) and X 2 ( z ) have ROCs R1 and R2 and inverse transforms x1[n] and x 2 [n ]
respectively. We see from Entries 5 and 6 of z-transform table, X1( z ) and X 2 ( z ) can each have
two possible inverse transforms

ìï (0.5)n u[n ], z > 0.5

x1[n ] = ïí

X1( z ) ¬¾¾

ïï-(0.5)n u[-n - 1], z < 0.5
ïî

ìï 2n u[n ], z >2
x2 [n ] = íï

X 2 ( z ) ¬¾¾

ïï-2n u[-n - 1], z <2
ïî

It thus follows there are 22 = 4 ways of forming x [n ] = x1[n ] + x2 [n ] , depending on R1 and R2

R1 R2 R1 Ç R2

1 z > 0.5 z >2 z >2

2 z > 0.5 z <2 0.5 < z < 2

3 z < 0.5 z >2 Æ

4 z < 0.5 z <2 z < 0.5

As can be seen, only Entries 2 and 3 satisfy R X Ê R1 Ç R2 . However, we must discard Entry 3 since
it leads to x [n ] = -(0.5)n u[-n - 1] + 2n u[n ] which does not have a z-transform. In other words, Entry
3 contradicts the problem statement, that is, there is an X(z ). Hence, from Entry 2

x [n ] = x1[n ] + x2 [n ] = (0.5)n u[n ] - 2n u[-n - 1]

Curtin University, Australia 2-22

YH Leung (2005, 2006, 2007, 2012, 2018)
6 Discrete-Time Systems

 Systems can be: continuous-time or discrete-time

linear or non-linear
time-invariant or time-varying
stable or unstable
causal or non-causal
with memory or memoryless
invertible or non-invertible

We focus on discrete-time linear time-invariant, stable and unstable, and

causal and non-causal systems

 Let  be a discrete-time system with input x[ n ] and output y[ n ]

y [n ] =  { x [n ]} (26)

x[n]  y[n]

(i)  is linear if, for any a, b Î 

 {ax1[n ] + bx2 [n ]} = a  { x1[n ]} + b  { x2 [n ]} (27)

(ii)  is time-invariant if, for all no Î 

 { x [n - no ]} = y [n - no ] (28)

(iii)  is stable in the bounded-input-bounded-output (BIBO) sense if and only if

for all input sequences satisfying

x [ n ] £ Bx < ¥ , "n (29)

we can find a By such that

y [n ] £ By < ¥ , "n (30)

(iv)  is causal if, for all no Î  , y [no ] depends only on { x [n ], n £ no }

Curtin University, Australia 2-23

YH Leung (2005, 2006, 2007, 2012, 2018)
Example 7

(a) y [n ] = x[n ] + x [n - 1]

Linearity
 {ax1[n ] + bx2 [n ]} = (ax1[n ] + bx2 [n ]) + (ax1[n - 1] + bx2 [n - 1])
= a ( x1[n ] + x1[n - 1]) + b ( x2 [n ] + x2 [n - 1])
= a { x1[n ]} + b { x2 [n ]}

\ System is linear

Time-Invariance
 { x[n - no ]} = x[n - no ] + x[n - 1- no ]

y [n - no ] = x [n - no ] + x[n - no - 1]

Clearly  { x[n - no ]} = y [n - no ]

\ System is time-invariant

(b) y [n ] = x[-n ]

Linearity
 {ax1[n ] + bx2 [n ]} = ax1[-n ] + bx2 [-n ]
= a { x1[n ]} + b { x2 [n ]}

\ System is linear

Time-Invariance
 { x[n - no ]} = x [-n - no ]

y [n - no ] = x [-(n - no )] = x [-n + no ]

\ System is not time-invariant

Curtin University, Australia 2-24

YH Leung (2005, 2006, 2007, 2012, 2018)
 Suppose  is linear time-invariant (LTI). Define

h[n ] =  {d [n ]} (31)

h[ n ] is unit sample response, or impulse response, of 

(i) It follows from Eqs. (18), (27) and (28) that

ìï ¥ üï ¥
y [n ] =  { x [n ]} =  ïí å x [k ] d[n - k ]ïý = å x [k ]  {d[n - k ]}
ïîïk =-¥ ïþï k =-¥
¥
= å x [k ] h[n - k ]
k =-¥

i.e. y [n ] = x [n ] * h[n ] = h[n ] * x [n ] (32)

Eq. (32) shows an LTI system is completely characterised by h[n ]

(ii) z-transform of Eq. (32) is given by

Y (z) = H (z) X (z) (33)

where H ( z ) =  {h[n ]} 4 (34)

Hence an LTI system is also completely characterised by H (z ). Indeed, the

system is often identified as H (z )

H (z ) is transfer function of the system. Also called system function

x[n] h[n] y[n] X(z) H(z) Y(z)

Fig. 3 Time- and transform-domain representations of a system

(iii) Frequency response of system is given by the DTFT

H (e j q ) = H ( z ) z = e j q (35)

4
Alternatively, since X ( z ) = 1 if x[n ] = d[n ] , so by Eq. (33), Y (z ) x[ n ]=d[n ] = H ( z) . But
Y ( z ) x [ n ]=d[n ] =  {y [n ] x [ n ]=d [ n ] } and by definition, y [n] x[ n ]=d[n ] = h[n] . Therefore, it follows that
 {h[n ]} = H ( z )

Curtin University, Australia 2-25

YH Leung (2005, 2006, 2007, 2012, 2018)
(iv) It can be shown an LTI system is stable if and only if5
¥
å h[n ] < ¥ (36)
n =-¥

¥ ¥
But å h[n] = å h[n ] z-n
z =1
n =-¥ n =-¥

Therefore, a stable H (z ) must include the unit circle in its ROC

(v) As can be seen from Eq. (32), an LTI system (stable or unstable) is causal if
and only if h[ n ] = 0 for n < 0

(vi) If  is LTI and causal, and x[ n ] is also causal, then Eq. (32) simplifies to
n n
y [n ] = å x[k ] h[n - k ] = å h[k ] x[n - k ] (37)
k =0 k =0

 Suppose an LTI system has impulse response h[n ] = {h0 , h1, h2 , h3 }3n =0 and
input x[n ] = {a, b, g }2n =0

Because of linearity, output response can be found by adding three impulse

response sequences, fired off at n = 0, n = 1 and n = 2, and scaled by a, b
and g respectively, as shown below

n=0 n =1 n=2 n=3 n=4 n=5

a ah0 ah1 ah2 ah3

b b h0 b h1 b h2 b h3
g g h0 g h1 g h2 g h3
y [n ] ah0 ah1 ah2 ah3 b h3 g h3
+b h0 +b h1 +b h2 +g h2
+g h0 +g h1

Observe that, skewing above table 45° clockwise yields precisely the con-
volution table of pp. 7, thus demonstrating principle of convolution table method

5
See A. V. Oppenheim and R. W. Schafer, Discrete-Time Signal Processing, 3rd ed., Prentice-Hall,
2010, pp. 59-60

Curtin University, Australia 2-26

YH Leung (2005, 2006, 2007, 2012, 2018)
7 Interconnected Systems

 It can be shown (exercise) from Eqs. (32) and (33) that:

Cascaded Systems Parallel Systems

h1[n] h2[n] h1[n]

h2[n] h1[n] h2[n]

h1[n] * h2[n] h1[n] + h2[n]

H1(z)H2(z) H1(z) + H2(z)

Fig. 4 System interconnections

Example 8

w[n]
x[n] A(z) B(z) y[n]

C(z)

Defining w[n] as shown and taking z-transforms of x[n], y[n], and w[n], we get, at summer output

W ( z ) = A( z ) X ( z ) + B( z ) C( z ) W ( z )

A( z )
\ W (z) = X (z)
1 - B( z ) C ( z )

But Y ( z ) = B( z ) W ( z )

A( z )
\ Y ( z ) = B( z ) X (z )
1 - B( z ) C ( z )

Y (z) A( z ) B( z )
Hence H (z) = =
X (z) 1 - B( z ) C ( z )

Curtin University, Australia 2-27

YH Leung (2005, 2006, 2007, 2012, 2018)
8 Linear Constant-Coefficient Difference Equations

 Linear constant-coefficient difference equations (LCCDEs) describe an

important class of discrete-time LTI systems6

 An N th order LCCDE is defined by

N M
å ak y [n - k ] = å bm x[n - m] (38)
k =0 m =0

where ak and bm are constants, and a0 normally equals 1

 Taking z-transform of Eq. (38), we get

N M
å ak Y (z) z-k = å bm X (z) z-m
k =0 m =0

M
å bm z-m
m =0
i.e. Y (z) = N
X (z) (39)
å ak z -k

k =0

Transfer function of LCCDE system (see Eq. (33)) given by

M M
bm -m
Y (z)
å bm z-m å a
z
m =0 m =0 0
H(z) = = N
= N
(40)
X (z) a
å ak z -k
1 + å k z-k
a
k =0 k =1 0

6
LCCDE systems form a special class of LTI systems. Can you think of an LTI system that cannot be
described by an LCCDE?

Curtin University, Australia 2-28

YH Leung (2005, 2006, 2007, 2012, 2018)
 Observe
M M M
å bm z -m
z N
å bm z M -m z N
å bM -q zq
m =0 m =0 q =0
H(z) = N
= N
= N
(41)
å ak z -k
z M
å ak z N -k
z M
å aN -p z p

k =0 k =0 p =0

ì
ï zerosüïï ìï numerator üï
ï
í ý of H(z) given by roots of ïí ïý polynomial in z
ï
ïpolesï
î ï
þ ïîïdenominator ïþï

Beware!
Although H(z) is often expressed as a ratio of two polynomials in z-1 , its poles
and zeros are found by expressing it as a ratio of two polynomials in z

Example 9
1
H (z) =
1 - az-1
It is often said H(z) has no zeros and only one pole at z = a. But writing H(z) as follows
1 z z
H (z) = -1
⋅ =
1 - az z z - a

we see H (z) actually also has a zero at z = 0

ìïN > M üï ì ü
 It follows from Eq. (41) that if ïí ïý , then H(z) has ïïíN - M zerosï ï
ý at z = 0
ïîïM > N ïþï ïîïM - N polesï
ï
þ

Curtin University, Australia 2-29

YH Leung (2005, 2006, 2007, 2012, 2018)
 A less obvious result for H(z) is

number of poles = number of zeros

Example 10
z-1 1
(a) H (z) = -1
=
1 - az z -a

Clearly, H (z) has a pole at z = a. But as z  ¥ , H ( z )  0 . Hence, H (z) also has a zero
at ¥

z2 - 1 ( z - 1)( z + 1)
(b) H ( z ) = 1 - z -2 = 2
=
z z2
H (z) has 2 poles, at z = 1, and also 2 zeros at z = 0

(c) H (z) = z

Clearly, H (z) has a zero at z = 0. Its pole though is at z  ¥ . (Note, with pole at ¥, H(z)
can only have ROC RH < ¥ , meaning it can only be non-causal.)

Curtin University, Australia 2-30

YH Leung (2005, 2006, 2007, 2012, 2018)
 Let x[n] be a known input sequence. Output sequence y[n] can be found by:

(a) solving LCCDE formally through its homogeneous and particular solutions

(b) using z-transforms

y [n ] =  -1 {Y ( z )} =  -1 {H ( z ) X ( z )} (42)

(c) determining the linear convolution

y [n ] = x [n ] * h[n ] = h[n ] * x [n ] (43)

(d) assuming y [no - 1], , y [no - N ] are known, compute the forward recursion
N M
a bm
y [n ] = - å k y [n - k ] + å x[n - m], n = no , no + 1,  (44)
k =1 a0 m =0 a0

Alternatively, assuming y [no ], , y [no - N + 1] are known, compute the back-

ward recursion
N -1 M
ak b
y [n - N ] = - å y [n - k ] + å am x[n - m], n = no , no - 1,  (45)
k =0 aN m =0 N

 Clearly, to solve Eqs. (42) and (43), RH must be known

In other words, knowing LCCDE is not enough – more information is required

 Therefore, for example, suppose LCCDE system is causal and stable

(i) h[n] causal  ROC of H(z) has form R- < z

(ii) h[n] stable  ROC includes unit circle

\ h[n] causal and stable  H(z) must have all its poles inside unit circle

Curtin University, Australia 2-31

YH Leung (2005, 2006, 2007, 2012, 2018)
Example 11

Consider LCCDE
y [n ] = a y [n - 1] + x [n ]

1
Clearly H (z) =
1 - az-1

Clearly, H(z) has a pole at z = a. Therefore, there are two possible ROCs: { z > a } and { z < a } ,
and from z-transform table

 1
h[n ] = a n u [n ] ¬¾¾
 H (z) = , z > a
1 - az-1

 1
h[n ] = - a n u [-n - 1] ¬¾¾
 H (z) = , z <a
1 - az-1

We show below how we can find both forms of h[n] recursively

Suppose y[-2] = 0 and x[n] = d [n]. By forward recursion

y [n ] = a y [n - 1] + x [n ]

N y[n - 1] x[n] y[n]

-2 0 0 0

-1 0 0 0

0 0 1 1
1 1 0 A
2
2 a 0 a

3 a2 0 a3

¥
Thus y [n ] = h[n ] = {0, 0, 1, a, a 2 , a3 , } = a n u[n ]
n=-2

For backward recursion, suppose y[2] = 0 and x[n] = d [n]. Re-arranging LCCDE as follows
1
y [n - 1] = ( y [n ] - x [n ])
a

N y[n] x[n] y[n - 1]

2 0 0 0

1 0 0 0

0 0 1 -1 a
2
-1 -1 a 0 -1 a
2 3
-2 -1 a 0 -1 a
3
-3 -1 a 0 -1 a
4

y [n ] = h[n ] = {, - a-3 , - a-2 , - a-1, 0, 0, 0}

2
Thus = - a n u[-n - 1]
n=-¥

Curtin University, Australia 2-32

YH Leung (2005, 2006, 2007, 2012, 2018)
9 Partial Fraction Expansion (PFE)

 Recall Eq. (40). Factorising H(z)

M M
å bk z-k b0
 (1 - ck z-1)
k =0 k =1
H (z) = N
= N
(46)
a0
å ak z -k
 (1 - dk z -1
)
k =0 k =1

we see: non-zero zeros given by c1, , cM

non-zero poles given by d1, , dN

 If d k1 = d k2 =  = d ks for some k1 ¹ k2 ¹  ¹ ks , then we say d k1 is an sth

order pole, or a repeated pole with multiplicity s

 Suppose H(z) has only one repeated pole (with multiplicity s), and poles are
so indexed that d1, , dN -s are simple poles while dN -s +1 is the repeated
pole

PFE of H(z) is given, with dN -s +1 =  = dN , by

M M

N(z)
å bk z-k b0  (1 - ck z-1)
k =0 k =1
H (z) = = N
= N
D( z )
å ak z-k a0  (1 - d k z-1)
k =0 k =1

N ( z ) M -N N -s
An s
Cm
H (z) = = å Br z-r + å -
+ å -1 m
(47)
D( z ) r =0 n =1 1 - d n z
1
m =1 (1 - dN -s +1z )

where residues Br , An and Cm are given by

N(z)
An = (1 - d n z-1) H ( z ) = N
(48)
z =d n
a0  (1 - d k z -1
)
k =1
k ¹n z =dn

Curtin University, Australia 2-33

YH Leung (2005, 2006, 2007, 2012, 2018)
 1 ìï d s -m -1 ù ï
ü
Cm = íï é(1 - d w )s
H (w ) ýï
ê N -s +1 ûúï
(s - m )!(-dN -s +1)s -m ïîï dw s -m ë þïw =dN--1 s +1

1 ìï d s -m é N (w -1) ùüï
Cm = ï ï ê úï
úïïý
s -m ï
(49)
(s - m )!(-dN -s +1) í dw s -m ê N -s
êa úï
ê 0  (1 - d k w ) úïï
ïï
ïîï ëê k =1 úïw =dN--1 s +1
ûþ

and Br are found by long division of N(z) by D(z) with division stopping
when degree of remainder is less than N

 Can readily extend above procedure to H(z)’s with more than one repeated
pole

 h[n] is found with following z-transforms

z-r = ¬¾¾

 = d[n - r ]
ìïanu[n ] 1
1 ï ( k -1)!
 = ïí

= ¬¾¾
1 - az-1 ïï-anu[-n - 1] 1
ïî ( k -1) !

ìï(n + 1) anu[n ] 1
1 ïï  ( k -1)!
= ¬¾¾
 = í
-1 2
(1 - az ) ïï-(n + 1) a u[-n - 2] 1
n
îï ( k -1)!

ìï 1 (n + 1)(n + 2) anu[n ] 1
1 ïï 2 ( k -1)!
= ¬¾¾
 = í
(1 - az-1)3 ïï- 1 (n + 1)(n + 2) anu[-n - 3] 1
îï 2 ( k -1)!

ìï 1 (n + 1)(n + 2)  (n + k - 1) anu[n ]
1 ï ( k -1)!
 = ïí

= ¬¾¾
(1 - az-1)k ïïï- ( k -11)! (n + 1)(n + 2)  (n + k - 1) anu[-n - k ]
î

 Since it is necessary that h[n ]  0 as n  ¥ for H(z) to be stable, above

z-transform pairs show causal poles of a stable H(z) must lie strictly inside
unit circle and anti-causal poles must lie strictly outside unit circle7

7
This is consistent with a result derived in pp. 26 where it is shown ROC of a stable LTI system must
include the circle unit

Curtin University, Australia 2-34

YH Leung (2005, 2006, 2007, 2012, 2018)
Example 12

8 - 7.5z-1 + 4.03z-2 - 0.926z-3 + 0.174z-4 - 0.018z-5

H (z) =
(1 - (0.2 + j 0.4)z-1)(1 - (0.2 - j 0.4)z-1)(1 - 0.3z-1)
2

8 - 7.5z-1 + 4.03z-2 - 0.926z-3 + 0.174z-4 - 0.018z-5

=
1 - z-1 + 0.53z-2 - 0.156z-3 + 0.018z-4

H (z) has M = 5 non-zero zeros, N = 4 non-zero poles and M - N = 1 pole at z = 0. Two of the
non-zero poles form a complex conjugate pair at d1,2 = 0.2  j 0.4 , and the other non-zero pole is
at d3 = 0.3 but with a multiplicity of s = 2. Thus PFE will have form

M -N N -s s
An Cm
H (z ) = å Br z-r + å 1- d -1
+ å (1 - d -1 m
r =0 n=1 nz m=1 N -s +1z )
1 2 2
An C
= å Br z-r +å -1
+ å (1 - 0.3mz-1)m
r =0 n=11 - d n z m=1

(i) B0 and B1 found as follows.

-z-1 + 1
0.018z-4 - 0.156z-3 + 0.53z-2 - z-1 + 1 -0.018z-5 + 0.174z-4 - 0.926z-3 + 4.03z-2 - 7.5z-1 + 8
-0.018z-5 + 0.156z-4 - 0.53z-3 + z-2 - z-1
0.018z-4 - 0.396z-3 + 3.03z-2 - 6.5z-1 + 8
0.018z-4 - 0.156z-3 + 0.53z-2 - z-1 + 1
- 0.24z-3 + 2.5z-2 - 5.5z-1 + 7

Hence B1 = -1 and B0 = 1

(ii) A1 and A2 found as follows.

A2 = (1 - d 2 z-1 ) H ( z )
z=d2
-1
8 - 7.5z + 4.03z-2 - 0.926z-3 + 0.174z-4 - 0.018z-5
= = 1 - j2
(1 - (0.2 + j 0.4)z-1)(1 - 0.3z-1)
2
z=0.2- j 0.4

A1 = A2* = 1 + j 2

(iii) C1 and C2 found as follows.

ìïï d 2-2 ü
C2 =
1
í é(1 - d w )2 H (w -1 )ù ýïï , d3 = 0.3
ë û
(2 - 2)!(-d3 )2-2 ïîï dw 2-2 ïþïw =1 d
3
3

ì 8 - 7.5w + 4.03w 2 - 0.926w 3 + 0.174w 4 - 0.018w 5 ïüï

ï
= (1) ⋅ ï
í ý
ï
ï
î (1 - (0.2 + j 0.4)w )(1 - (0.2 - j 0.4)w ) ïþï
w =1 0.3
= 4

Curtin University, Australia 2-35

YH Leung (2005, 2006, 2007, 2012, 2018)
 ïìï d 2-1 é ü
ï
(1 - d3w )2 H (w -1 )ûù ï
1
C1 = í
2-1 ï 2-1 ë
ý , d3 = 0.3
(2 - 1)!(-d3 ) îï dw ï
ïw =1 d3
þ
ìï d é 2 3 4 5 ùüï
=
1
⋅ íï ê 8 - 7.5w + 4.03w - 0.926w + 0.174w - 0.018w úýï
(-0.3) ï ê (1 - (0.2 + j 0.4)w )(1 - (0.2 - j 0.4)w ) ú ï
î dw
ï ë ûþïw =1 0.3
1 ìï d é 8 - 7.5w + 4.03w 2 - 0.926w 3 + 0.174w 4 - 0.018w 5 ùüï
= ⋅ïí ê
ê
úï
úý
(-0.3) ïî dw
ï ë 1 - 0.4w + 0.2w 2 ï
ûþïw =1 0.3
1 ìï 8 - 7.5w + 4.03w 2 - 0.926w 3 + 0.174w 4 - 0.018w 5
= ⋅ ïí (-1)(-0.4 + 0.4w )
(-0.3) ïïî (1 - 0.4w + 0.2w 2 )2
ü
- 7.5 + 8.06w - 2.778w 2 + 0.696w 3 - 0.09w 4 ï
+ ï
ý
1 - 0.4w + 0.2w 2 ï
ï
þw =1 0.3
=1

Thus
1+ j 2 1- j 2 1 4
H ( z ) = 1 + (-1) ⋅ z-1 + + + +
1 - (0.2 + j 0.4)z-1 1 - (0.2 - j 0.4)z-1 1 - 0.3 z-1 (1 - 0.3z-1 )2

Suppose H(z) is causal. Then, impulse response is given by8

h[n ] = d[n ] - d[n - 1]

+ (1 + j 2)(0.2 + j 0.4)n u[n ] + (1- j 2)(0.2 - j 0.4)n u[n ]
+ (0.3)n u[n ] + 4(n + 1)(0.3)n u[n ]

= d[n ] - d[n - 1] + 2Re {(1 + j 2)(0.2 + j 0.4)n } u[n ] + (4n + 5)(0.3)n u[n ]

Exercise: Write a MATLAB program to evaluate first 10 terms of h [n].

 PFE procedure is implemented in MATLAB by the function residuez (in

signal processing toolbox)

8
Alternatively, since
1+ j 2 1- j 2 2 - 2z-1
+ =
1 - (0.2 + j 0.4)w 1 - (0.2 - j 0.4)w 1 - 0.4z-1 + 0.2z-2
it follows (exercise) from entries 11 and 12 of the z-transform tables, that
ïì 2 - 2 z-1 ïüï n
 -1 ïí ý = 2 ( 0.2 ) (cos qo n - 2 sin qo n) u[n ], qo = cos
-1
0.2
ïï1 - 0.4z + 0.2z ïï
-1 -2
î þ
It can be readily verified using MATLAB that above sequence equals 2Re {(1 + j 2)(0.2 + j 0.4)n } u [n ]

Curtin University, Australia 2-36

YH Leung (2005, 2006, 2007, 2012, 2018)
 If H(z) is causal, h[n] can also be found by performing a long division of N(z)
by D(z) for as many terms as we wish
However, in above procedure, round-off errors will accumulate as n increases
and computed h[n] will not be accurate for n large

Example 12 (revisited)

8 - 7.5z-1 + 4.03z-2 - 0.926z-3 + 0.174z-4 - 0.018z-5

H (z) =
(1 - (0.2 + j 0.4)z-1)(1 - (0.2 - j 0.4)z-1)(1 - 0.3z-1)
2

8 - 7.5z-1 + 4.03z-2 - 0.926z-3 + 0.174z-4 - 0.018z-5

=
1 - z-1 + 0.53z-2 - 0.156z-3 + 0.018z-4

We demonstrate here a method to find the causal h[n] by long division. It is instructive to compare and
contrast long division shown below with that shown in pp. 35 to find the residues B0 and B1

8 + 0.5z-1 + 0.29z-2 + 0.347z-3 

-1 -2 -3 -4
1- z + 0.53z - 0.156z + 0.018 z 8 - 7.5 z-1 + 4.03 z-2 - 0.926 z-3 + 0.174 z-4 - 0.018 z-5
8 - 8z-1 + 4.24z-2 - 1.248z-3 + 0.144z-4
0.5z-1 - 0.21z-2 + 0.322z-3 + 0.03z-4 - 0.018z-5
0.5z-1 - 0.5z-2 + 0.265z-3 - 0.078z-4 + 0.009z-5
0.29z-2 + 0.057z-3 + 0.108z-4 - 0.027z-5
0.29z-2 - 0.29z-3 + 0.1537z-4 - 0.0452z-5 + 0.0052z-6
0.347z-3 - 0.0457z-4 + 0.0182z-5 - 0.0052z-6
0.347z-3 - 0.347z-4 

Quotient gives first 4 terms of h[n], namely, {8, 0.5, 0.29, 0.347, }

Curtin University, Australia 2-37

YH Leung (2005, 2006, 2007, 2012, 2018)
Appendix I – Proof of z-Transform Time-Shifting Property

 Consider the sequence y [n ] = x [n - no ] . It follows from Eq. (24) that

¥ ¥ ¥
Y (z) = å y [n ] z-n = å x[n - no ] z-n = å x[m] z-( m+no )
n=-¥ n=-¥ m=-¥
¥
= z-no å x[m] z-m = z-no X ( z )
m=-¥

Following discussion relates to ROC of y[n]

 Suppose x [n ] = a n u[n ] with X ( z ) = 1 (1- az-1 ) , R X = { a < z } (z-transform table Entry 5)

If no = -1 , i.e. if x[n] shifted one sample left, then z-no = z and Y(z) will not converge as
z  ¥ . Thus, RY = R X { z  ¥} = { a < z < ¥} , i.e. RY is R X with ¥ deleted

But if no = 2 , i.e. if x[n] shifted 2 samples right, then z-no = z-2 and Y(z) will not converge
at z = 0. But R X already has form { a < z } which excludes z = 0. Thus, RY = R X and there
is no change in ROC

 Suppose next x [n ] = -a n u[-n - 1] with X ( z ) = 1 (1- az-1 ) , R X = { z < a } (z-transform table

Entry 6)

If no = -1 , i.e. if x[n] shifted one sample left, then z-no = z and Y(z) will not converge as
z  ¥ . But R X already excludes z  ¥ . Therefore, RY = R X and there is no change in
ROC

If no = 2 , i.e. if x[n] shifted 2 samples right, then z-no = z-2 and Y(z) will not converge at
z = 0. Thus, RY = RX {z = 0} = {0 < z < a } , i.e. RY is R X with origin deleted

 Finally, reversing above arguments, we see that, starting with a sequence such as
x [n ] = a n +1u[n + 1] whose ROC is R X = { a < z < ¥} (see above), shifting x[n] one
sample right will result in y [n ] = x [n - 1] = a n u [n ] whose ROC is RY = { a < z } . In other
words, RY is R X with ¥ added. Likewise, shifting x [n ] = -a n-2u[-n + 1] (whose ROC is
R X = {0 < z < a } ) 2 samples left will result in origin being added to the ROC

Curtin University, Australia 2-38

YH Leung (2005, 2006, 2007, 2012, 2018)
(a) x [n ] = (-0.7)n u[n ]
...
−2 −1 0 1 2 3 4 n

From Entry 5 of z-transform table,

1
X (z) = , z > -0.7 = 0.7
1 + 0.7z-1

Above ROC is consistent with Entry 4 of ROC chart

(b) x [n ] = (-0.7)n u[n - 1]

...
−2 −1 0 1 2 3 4 n

x [n ] = (-0.7)(-0.7)n-1 u[n - 1]

Therefore, from Entry 5 of z-transform table, and time-shift and linearity properties

z -1
X ( z ) = (-0.7)
1 + 0.7 z-1

As for R X , according to proof of time-shifting property, there should be no change, i.e.

R X = { z > 0.7} . This is consistent with Entry 4 of ROC chart

...
(c) x [n ] = (-0.7)n u[n + 1] n
−2 −1 0 1 2 3 4

x [n ] = (-0.7)-1 d[n + 1] + (-0.7)n u[n ]

Therefore, from Entries 4 and 5 of z-transform table, and time-shift and linearity properties

1 1
X (z) = - z +
0.7 1 + 0.7z-1

and R X = {all z except z  ¥} Ç { z > -0.7 } = {0.7 < z < ¥}

Above ROC is consistent with discussion in proof of time-shifting property, and with Entries 2
and 4 of ROC chart

Curtin University, Australia 2-39

YH Leung (2005, 2006, 2007, 2012, 2018)
Appendix II – ROCs in Linearity Property

 We first consider following 7 examples

(i) x [n ] = x1[n ] + x2 [n ] = a n u[n ] - b n u [n - 4], a ¹ b

Then by z-transform table Entry 5 and time-shifting property

1 b 4 z-4
X ( z ) = X1( z ) + X 2 ( z ) = -
1 - az-1 1 - bz-1

where for X1( z ) and X 2 ( z ) to converge, we require z > a and z > b respectively. In
other words, for X(z) to converge, we require z to satisfy, simultaneously, z > a and
z > b . In other words
R X = R1 Ç R2

where R1 = { z > a } and R2 = { z > b }

(ii) Suppose now a = b, i.e.

x [n ] = x1[n ] + x2 [n ] = a n u[n ] - a n u[n - 4]

Since R1 = R2 = { z > a } , it would now appear

R X = R1 Ç R2 = R1 Ç R1 = R1 = { z > a }

But

1 a 4 z-4 1 - a 4 z-4
X ( z ) = X1( z ) + X 2 ( z ) = - =
1 - az-1 1 - az-1 1 - az-1

(1 - az-1 )(1 + az-1 + a 2 z-2 + a3 z-3 )

=
1 - az-1

= 1 + az-1 + a 2 z-2 + a3 z-3

which corresponds to the finite length sequence {1, a, a 2 , a3 }

3
. Therefore, by ROC chart
n =0
Entry 1
R X = { z > 0}

In other words, because of pole-zero cancellation at z = a, RX has expanded resulting in

R X É R1 Ç R2

Curtin University, Australia 2-40

YH Leung (2005, 2006, 2007, 2012, 2018)
(iii) Consider next

x [n ] = x1[n ] + x2 [n ] = - a n u [-n - 1] - a n u[n - 4]

It would appear z-transform is given by

1 a 4 z- 4
X ( z ) = X1( z ) + X 2 ( z ) = - = 1 + az-1 + a 2 z-2 + a3 z-3
1 - az-1 1 - az-1

as in Example (ii). But for X1( z ) and X 2 ( z ) to exist, we require respectively z < a and
z >a which contradicts one another. In other words, X(z) cannot exist. As for RX
(assuming it is meaningful to speak of one!), it is given by

RX = { z < a } Ç { z > a } = Æ

With reference to the linearity property, we see

R1 Ç R2 = Æ = R X

(iv) We next extend previous example (iii)

x [n ] = x1[n ] + x2 [n ] = (-a n u [-n - 1] - a n u [n ]) + b n u [n ], b ¹ a

Clearly, for X1( z ) to converge, we require z < a and z > a meaning R1 = Æ ; while for
X 2 ( z ) to converge, we require z > b , i.e. R2 = { z > b } . In other words, for X(z ) to
converge, three conditions must be satisfied, and because of the contradiction between the first
two conditions, R X = Æ

With reference to the linearity property, we see

R1 Ç R2 = Æ Ç { z > b } = Æ = R X

(v) Suppose b = a in Example (iv), i.e.

x [n ] = x1[n ] + x2 [n ] = (-a n u [-n - 1] - a n u [n ]) + a n u [n ]

Clearly x [n ] = - a n u[-n - 1]

which from Entry 6 of z-transform table, has transform

1
X (z ) = , z < a
1 - az-1

In this case, we see

R1 Ç R2 = Æ Ç { z > a } = Æ

while R X = { z < a } É R1 Ç R2

Curtin University, Australia 2-41

YH Leung (2005, 2006, 2007, 2012, 2018)
(vi) Consider next

x [n ] = x1[n ] + x2 [n ] = (-a n u [-n - 1] - b n u [n ]) + b n u [n ], b < a

Clearly, R1 = { b < z < a } and R2 = { b < z } . Accordingly,

R1 Ç R2 = { b < z < a } Ç { b < z } = { b < z < a }

But x[n] can be simplified as follows

x [n ] = - a n u[-n - 1]

whose ROC is
RX = { z < a }

Hence R X É R1 Ç R2

(vii) Finally, consider

x [n ] = x0 [n ] + x1[n ] + x2 [n ] + x3 [n ] +  = a0 d [n ] + a1d[n - 1] + a 2d [n - 2] + a3 d [n - 3] + 

Taking z-transforms, we get from the linearity and time-shifting properties

X ( z ) = X 0 ( z ) + X1( z ) + X 2 ( z ) + X 3 ( z ) +  = 1 + a1z-1 + a 2 z-2 + a3 z-3 + 

?
with R0 Ç R1 Ç R2 Ç R3 Ç  = {all z} Ç { z > 0} Ç { z > 0} Ç  = { z > 0}

Now, x[n] as written above is simply sample-by-sample decomposition of a n u[n ] . See Eq.
(18). Accordingly, from Entry 5 of z-transform table

RX = { z > a }

Difficulty here is above results appear to violate the condition

R X Ê (R0 Ç R1 Ç R2 Ç R3 Ç )

Paradox resolved by noting linearity property applies only to a finite sum of subsequences. For
infinite sums, one has to study behaviour of the sum in the limit

Curtin University, Australia 2-42

YH Leung (2005, 2006, 2007, 2012, 2018)
 Lessons

1. In most cases, R X = R1 Ç R2 . See Examples (i), (iii) and (iv)

2. In some cases, R X can grow, i.e. R X É R1 Ç R2 , due to pole-zero cancellation

as in Example (ii), or sequence cancellation as in Examples (v) and (vi)

3. Linearity property applies only to finite sum of subsequences. See Example (vii)

4. Generally, a sketch of x[n] can be helpful before one attempts to find X(z).
See Examples (ii), (v), (vi) and (vii)

We next consider the inverse problem, i.e., find x[n] given X(z) and R X

5. If R X = Æ , then X(z) cannot be specified and no unique x[n] can be found.

Indeed, from Examples (iii) and (iv), we see there is an infinite number of
sequences with R X = Æ

6. From Examples (v) and (vi), we see that given a non-empty R X and an X(z),
the inverse transform is not unique. However, all these inverse transforms are
identical algebraically

7. R1 Ç R2 = Æ does not imply an inverse transform cannot be found since

R X may not be empty. See Example (v)

8. In Tutorial Problem 19, it is shown that if X(z) has partial fraction expansion

a1 aN
X ( z ) = X1( z ) +  + X N ( z ) = -1
++ (50)
1 - b1z 1 - bN z-1

where a1, , aN ¹ 0 (51)

and b1 ¹  ¹ bN ¹ 0 (52)

then X(z) will not exhibit any pole-zero cancellations

Curtin University, Australia 2-43

YH Leung (2005, 2006, 2007, 2012, 2018)
Problems

1. Consider the continuous-time signal

xc (t ) = A cos (8pt )

(a) What is the (fundamental) period of xc (t ) ?

(b) Suppose xc (t ) is being sampled with a sampling period of T = 1 16 sec to yield
the discrete-time sequence x[n] = xc (nT ) . Determine the fundamental period N
of x[n], and comment on the value NT.

(c) Repeat Part (b) for T = 1 18 sec and T = 1 401 sec .

M
2. Suppose x = { x[n]} 2 , y = { y[n]}N 2 , and w = x * y = {w[n]}K 2 .
n=M1 n= N1 n= K1

Show that K1 = M1 + N1

and K2 = M 2 + N2

3. Given x[n] = {3, -1, 2, 0, 4,1, - 2}4n=-2

and y[n] = {2, 1, 0, -1, 3}3n=-1

Find the sequences

(a) w1 = 2 x - y
(b) w2 = x ⋅ y
(c) w3 = x * y

4. Use the properties of DTFT and the table of DTFT pairs to find the DTFT of the
sequence
x[n] = a n , a <1

(This problem was solved in pp. 10 but the solution was based on the geometric series
formula Eq. (23).)

Curtin University, Australia 2-44

YH Leung (2005, 2006, 2007, 2012, 2018)
5. Consider the discrete-time signal

ìï{-1, 0, 1, 2, 1, 0, 1, 2, 1, 0, -1} , -3 £ n £ 7
x[n] = ïí
ïïî0, otherwise

Without explicitly evaluating the DTFT X (e jq ) , find

(a) X ( e jq )
q=0

(b) X (e jq )
q=p

(c) arg X (e jq )
p jq
(d) ò-p X (e ) dq

(e) Determine the signal whose DTFT is X (e- jq )

(f) Determine the signal whose DTFT is Re { X (e jq )}

6. Prove Parseval’s (Plancherel’s) Relation

¥
1
å x[n] y* [n] =
2p ò 2 p
X ( e jq ) Y * ( e jq ) d q
n=-¥

Hint. Consider the DTFT of g[n] = x[ n] * y* [-n]

7. The frequency response of a discrete-time linear time-invariant system is known

to satisfy
H (e jq ) = H (e j (q-p ) )
Determine h[5] where h[n] is the impulse response of the system.

8. A discrete-time linear time invariant system was subjected to the input sequence
x[ n] = ( 1 4 ) n u[n] . The output sequence was observed to have the form y[n] =
ad[ n] + bd[n -1] where a and b are real constants. Determine a and b, and the
impulse response h[n] of the system, given it is known that H (e j p 2 ) = 1 .

Curtin University, Australia 2-45

YH Leung (2005, 2006, 2007, 2012, 2018)
9. Determine the z-transform, including the ROC, for each of the following sequences. Use
the properties of z-transforms, the table of z-transform pairs, and the ROC chart.

x[n] = ( 12 ) u[n]
n
(a)

( 12 )
n
(b) x[n] = (u[n] - u[n -10])

(c) x[n] = - 2-n u[-n -1]

x[n] = (- 12 ) u[-n -1] + ( 12 )

n n
(d) u[n]

(e) x[n] = a n , a <1

(f) x[n] = a n , a >1

10. Determine the inverse z-transform for each of the following using the properties of
z-transforms, a table of z-transform pairs, the ROC chart, and the PFE.

1- 1 z -1
(a) X ( z) = 2
, 1 < z < 1
1+ 3 z -1 + 1 z -2 4 2
4 8

1- 1 z-1
(b) X ( z) = 2
, z > 1
1+ 3 z-1 + 1 z-2 2
4 8

11. Two LTI systems, one with impulse response h[n] and the other with impulse response
g[n], are cascaded together as shown below.

x[n] w[n] y[n]

h[n] g[n]

Show, using the definition of convolution, that the cascaded system can be represented
by a single system with impulse response h[n]*g[n].

x[n] y[n]
h[n] * g[n]

Curtin University, Australia 2-46

YH Leung (2005, 2006, 2007, 2012, 2018)
12. Are the following systems linear? Are they time-invariant?

(a)  { x[n]} = g[n] x[n] with g[n] given

åk=n
n
(b)  { x[n]} = x[k ]
o

(c)  { x[n]} = e x[ n]

(d)  { x[n]} = ax[n] + b

(e)  { x[n]} = x[-n + 1]

13. Example 3.12 of Oppenheim and Schafer, pp. 152, gives the transfer function of a
discrete-time LTI system that cannot be described a finite order LCCDE. Specifically,
H ( z ) = ln(1 + az-1 ) . Give other examples.

14. When the input to an LTI system is

( 12 )
n
x[n] = u[n] + 2n u[-n -1]

the output is

y[ n ] = 6 ( 12 ) u[ n ] - 6 ( 34 ) u[ n ]
n n

(a) Find the transfer function H(z) of the system. Where are its poles and zeros and
what is its ROC?
(b) Find the impulse response of the system for all n.
(c) Write the LCCDE that characterises the system.
(d) Is the system stable? Is it causal?

15. A causal LTI system is described by the LCCDE

y[ n] = 3
2
y[ n -1] + y[ n - 2] + x[ n -1]

(a) Find the system function H ( z ) = Y ( z ) X ( z ) and its ROC.

(b) Find the impulse response of the system.
(c) You should have found the system is unstable. Find a stable (non-causal) impulse
response that satisfies the difference equation.

Curtin University, Australia 2-47

YH Leung (2005, 2006, 2007, 2012, 2018)
16. The transfer function of an LTI system has the pole-zero plot shown below. Specify
whether each of the following statements is true, false, or cannot be determined from
the information given.
Im

0 Re

z-plane Unit circle

(a) System is stable.

(b) System is causal.
(c) If system is causal, then it must be stable.
(d) If system is stable, then it must have a two-sided impulse response.

17. A causal LTI system has transfer function

1 + z -1
H ( z) =
(1 - 1 z-1 )(1 + 1 z -1 )
2 4

(a) What is the region of convergence (ROC) of H(z)?

(b) Is the system stable? Explain.
(c) Find the z-transform of the input x[n], including its ROC, that yields the output

y[ n] = - 13 (- 14 ) u[n] - 43 (2)n u[-n -1]

n

Hint. First find the z-transform of y[n], including its ROC.

18. Given
 1
a nu[n] = ¬¾ =
1 - az-1
Use the properties of z-transform to find the z-transform of the following sequences.

(a) ( n + 1)a n u[n]

(b) (n + 1)(n + 2)a nu[n]

(c) (n + 1)(n + 2)  (n + k -1) a nu[n]

Curtin University, Australia 2-48

YH Leung (2005, 2006, 2007, 2012, 2018)
19. Suppose
a1 aN
H ( z ) = H1 ( z ) +  + H N ( z ) = -1
++
1 - b1 z 1 - bN z-1

where a1 ,  , a N ¹ 0

and b1 ¹  ¹ bN ¹ 0

Prove that H(z) cannot exhibit any pole-zero cancellations.

Curtin University, Australia 2-49

YH Leung (2005, 2006, 2007, 2012, 2018)
 Answers/Solutions

1. (a) Fundamental period is T p = 1 4

(b) N = 4, NT = T p
(c) For T = 1 18 , N = 9 and NT = 2T p . For T = 1 401 , sequence is not periodic.

2. Consider the convolution sum w[n] = å x[k ] y[n - k ] and the following set of figures.
k

x[k ]

0 M1 M2 k

y[k ]

N1 0 N2 k

y[-k ]

-N 2 0 -N1 k

y[-k + n] = y[n - k ]

n increasing

-N 2 + n 0 n -N1 + n k

As can be seen, as n increases from -¥ , x[k] and y[n-k] will start to overlap when

-N1 + n = M1

i.e. n = M1 + N1

and stop overlapping when

-N 2 + n = M 2

i.e. n = M 2 + N2

The starting index of w[n] is thus M1 + N1 , and its ending index is M 2 + N 2 .

Curtin University, Australia 2-50

YH Leung (2005, 2006, 2007, 2012, 2018)
3. (a) 2 x - y = {6, - 4, 3, 0, 9, -1, - 4}4n=-2
(b) x ⋅ y = {-2, 2, 0, - 4, 3}3n=-1
(c) x * y = {6,1, 3, -1,18,1, 3, - 6,11, 5, - 6}7n=-3

1 - a2
4. DTFT {a n
} = 1 + a 2 - 2a cos q

5. (a) 6
(b) 2
(c) -2q
(d) 4p
(e) {-1, 0, 1, 2, 1, 0, 1, 2, 1, 0, -1}3n=-7
{- 12 , 0, 12 , 1, 0, 0, 1, 2, 1, 0, 0, 1, 12 , 0, - 12 }n=-7
7
(f)

6. From the DTFT conjugation and time reversal properties

DTFT { y*[-n]} = Y * (e jq )

\ DTFT { x[n] * y*[-n]} = X (e jq )Y * (e jq )

Define g[ n] = x[n] * y* [-n]

It now follows from the definition of IDTFT that

1
g[ n] = ò X (e jq ) Y * (e jq ) e+ jnq d q
2p 2 p

whereupon
¥ ¥
1
g[0] =
2p ò 2 p
X ( e jq ) Y * ( e jq ) d q = å x[k ] y* [k - n] = å x[k ] y* [k ]
k =-¥ n=0 k =-¥

7. h[5] = 0

8. a= 16
17
,b= 4
17
, h[n] = 16 d[ n]
17
- 1 d[ n - 2]
17

Curtin University, Australia 2-51

YH Leung (2005, 2006, 2007, 2012, 2018)
 1
9. (a) , z > 1
1- 1 z -1 2
2

1 - ( 12 ) z-10
10
= 1 + 12 z-1 +  + ( 12 ) z-9 , z > 0
9
(b)
1- 1 z-1
2

1
(c) , z < 1
1- 1 z-1 2
2

(d) Transform does not exist.

1 - a2
(e) , a < z < 1
(1 + a 2 ) - a ( z + z-1 ) a

(f) Transform does not exist.

-4 (- 12 ) u[-n -1] - 3(- 14 ) u[n]

n n
10. (a)

x[ n] = 4 (- 12 ) u[ n ] - 3 (- 14 ) u[ n ]
n n
(b)

11. From the definition of the convolution sum

¥
w[ n] = h[n] * x[ n] = å h[k ] x[n - k ]
k =-¥
¥
and y[n] = g[n] * w[n] = å g[l ] w[n - l ]
l =-¥

Substituting the first equation into the second, we get

¥ ¥
y[n] = å g[l ] å h[k ] x[n - l - k ]
l =-¥ k =-¥

Next, we apply the change of index l + k = m so that

¥ ¥
y[n] = å g[l ] å h[m - l ] x[n - m]
l =-¥ m=-¥
¥ æ ¥ ö
= å ççççè å g[l ] h[m - l ]÷÷÷ x[n - m]
m=-¥ l =-¥
÷ø
¥
= å ( g[m] * h[m]) x[n - m]
m=-¥
= ( g[n] * h[n]) * x[n]

Curtin University, Australia 2-52

YH Leung (2005, 2006, 2007, 2012, 2018)
12.
Linearity Time-Invariance
(a)  
(b)  
(c)  
(d)  
(e)  

(a) Linear since

 {ax1[n] + bx2 [n]} = g[n](ax1[n] + bx2 [n])
= a g[n] x1[n] + b g[n] x2 [n]
= a { x1[n]} + b { x2 [n]}

Not time invariant since

 { x[n - no ]} = g[n] x[n - no ]

while y[ n - no ] = g[ n - no ] x[n - no ]

(d) Not linear since

 {a x[n]} = a (a x[n]) + b
¹ a { x[ n]} = a (ax[ n] + b)

Time invariant since

 { x[n - no ]} = ax[n - no ] + b = y[n - no ]

-1 1 -2 1 -3
13. H ( z) = ez = 1 + z -1 + z + z + , z >0
2! 3!

1 - 2 z -1
14. (a) H ( z) = , z > 3
1- 3 z -1 4
4
n-1
( 34 ) u[ n ] - 2 ( 34 )
n
(b) h[ n ] = u[ n -1]

(c) y[n] = ( 34 ) y[n -1] + x[n] - 2 x[n -1]

(d) System is stable and causal.

Curtin University, Australia 2-53

YH Leung (2005, 2006, 2007, 2012, 2018)
 z -1
15. (a) H ( z) = , z >2
1- 3 z -1 - z -2
2

( 52 ) 2n u[n] - ( 52 )(- 12 )
n
(b) h[ n ] = u[ n ]

h[ n ] = - ( 52 ) 2 n u[-n -1] - ( 52 )(- 12 )

n
(c) u[ n ]

16. (a) Cannot be determined

(b) Cannot be determined
(c) False
(d) True

17. (a) z >12

(b) H(z) is stable

1- 1 z -1
(c) X ( z) = 2
-1
1 - 2z

1
18. (a)
(1 - az-1 ) 2
2
(b)
(1 - az-1 )3
(k -1)!
(c)
(1 - az-1 )k

Curtin University, Australia 2-54

YH Leung (2005, 2006, 2007, 2012, 2018)
 a1 aN
19. H ( z) = -1
++
1 - b1 z 1 - bN z-1

It follows H(z) can be written, for appropriate values of c1 , , cN -1 , as follows

c1 + c2 z-1 +  + cN -1 z-( N -1)

H ( z) =
(1 - b1 z-1 )(1 - b2 z-1 )  (1 - bN z-1 )

Suppose the numerator can be factored such that, with no loss of generality, H(z) has
pole-zero cancellation at z = b1 . In other words, we can write H(z) as follows.

c1 (1 - b1 z-1 )(1 + d1 z-1 +  + d N -2 z-( N -2) )

H ( z) =
(1 - b1 z-1 )(1 - b2 z-1 )  (1 - b3 z-1 )

Given the above expression for H(z), suppose we now wish to perform a PFE. The
residue a1 is then given by

c1 (1 - b1 z-1 )(1 + d1 z-1 +  + d N -2 z-( N -2) )

a1 = (1 - b1z-1 ) X ( z ) = =0
z=b1 (1 - b2 z-1 )  (1 - b3 z-1 ) z=b1

which contradicts the original assumption that a1 ¹ 0 . In other words, H(z) cannot
exhibit any pole-zero cancellations.

Curtin University, Australia 2-55

YH Leung (2005, 2006, 2007, 2012, 2018)