Z-transform

Lecture 1

• Definitions of Z-transform,
• Unit circle
• convergence and ROC
• Relation between z-transform and DTFT,
• Z-transform and ROC (single sided and double sided):
• Numerical problem and Solution

Introduction

• Z-transform is a mathematical tool that uses to analyze the signals and systems in
complex z-plane.
• Z-transform is applicable only in LTI discrete time system.
• It is use to determine system function, impulse response, pole-zero and stability of the
discrete time systems.
• For frequency domain analysis there has a tool DTFT but it is not applicable on the
signals which are not absolutely summable. In that cases we can apply z-transform
technique for signal analysis and it represented in z-plane in a particular region where
z-transform is converge which is known as Region of convergence in z-plane.

• It is defined as by the equation,

∞

𝑋(𝑧) = ∑ 𝑥(𝑛) 𝑧 −𝑛
𝑛=−∞

• Region of Convergence (ROC):

Value of z for which 𝑋(𝑧) will be finite and the region in z-plane consist of such
points is known as Region of Convergence (ROC)

Application:

• Z- transform has many applications in physics and engineering such as analysis of

LTI systems, RADAR, astronomy, signal processing etc.
• Relation between DTFT and Z-transform:
• We know that DTFT of a signal x(n) as,
• 𝑋(𝑒 𝑗𝜔 ) = ∑∞
𝑛=−∞ 𝑥(𝑛) 𝑒
−𝑗𝜔𝑛

• And Z-transform of the signal x(n) is,

• 𝑋(𝑧) = ∑∞ 𝑛=−∞ 𝑥(𝑛) 𝑧
−𝑛

• By comparison we can write

• 𝑋(𝑧)|𝑧=𝑒 𝑗𝜔 = 𝑋(𝑒 𝑗𝜔 )
• So if we put 𝑧 = 𝑒 𝑗𝜔 in X(z), we will obtained DTFT of the signal, i.e., DTFT is
defined only on to the unit circle in z-plane
 • i) Z-transform and ROC of a finite duration and causal signal
• Let a causal finite duration signal 𝒙(𝒏) = {𝟏,↑ 𝟑, 𝟐, −𝟑, 𝟓}
• We know that z-transform of x(n) is,
• 𝑋(𝑧) = ∑∞ 𝑛=−∞ 𝑥(𝑛) 𝑧
−𝑛

• So for, 𝒙(𝒏) = {𝟏↑ , 𝟑, 𝟐, −𝟑, 𝟓} ;

4

𝑋(𝑧) = ∑ 𝑥(𝑛) 𝑧 −𝑛 = 𝑥(0)𝑧 −0 + 𝑥(1)𝑧 −1 + 𝑥(2)𝑧 −2 + 𝑥(3)𝑧 −3 + 𝑥(4)𝑧 −4

𝑛=0
= 1.1 + 3. 𝑧 −1 + 2. 𝑧 −2 + (−3). 𝑧 −3 + 5. 𝑧 −4
• ∴ 𝑋(𝑧) = 1 + 3𝑧 −1 + 2𝑧 −2 − 3𝑧 −3 + 5𝑧 −4
3 2
• 𝑋(𝑧) = 1 + 𝑧 + 𝑧 2 − 3𝑧 −3 + 5𝑧 −4
• Just observe that 𝑋(𝑧) given finite value for any value of z except 𝑧 = 0

So, for finite duration and causal signal, ROC is entire z plane except 𝒛 = 𝟎

ii) Z-transform and ROC of a finite duration and anti-causal signal

Lets an anti-causal finite duration signal 𝒙(𝒏) = { 𝟏, 𝟑, 𝟐, −𝟑, 𝟓, 𝟎

↑
}

We know that z-transform of x(n) is,

∞

𝑋(𝑧) = ∑ 𝑥(𝑛) 𝑧 −𝑛
𝑛=−∞
So for, (𝒏) = { 𝟏, 𝟑, 𝟐, −𝟑, 𝟓, 𝟎↑} ;
−1

𝑋(𝑧) = ∑ 𝑥(𝑛) 𝑧 −𝑛
𝑛=−5
= 𝑥(−5)𝑧 −(−5) + 𝑥(−4)𝑧 −(−4) + 𝑥(−3)𝑧 −(−3) + 𝑥(−2)𝑧 −(−2)
+ 𝑥(−1)𝑧 −(−1) = 1. 𝑧 5 + 3. 𝑧 4 + 2. 𝑧 3 + (−3). 𝑧 2 + 5. 𝑧1
∴ 𝑋(𝑧) = 𝑧 5 + 3𝑧 4 + 2𝑧 3 − 3𝑧 2 + 5𝑧

From the above equation we observe that 𝑋(𝑧) given finite value for any value of z except
𝑧=∞

So, for finite duration and anti-causal signal, ROC is entire z plane except 𝒛 = ∞

iii) Z-transform and ROC of a finite duration and non-causal signal

Let a non-causal finite duration signal 𝒙(𝒏) = {𝟏, 𝟑, 𝟐, −𝟑, 𝟓

↑
}

We know that z-transform of x(n) is,

 ∞

𝑋(𝑧) = ∑ 𝑥(𝑛) 𝑧 −𝑛
𝑛=−∞
So for, 𝒙(𝒏) = {𝟏, 𝟑, 𝟐,↑ −𝟑, 𝟓} ;
2

𝑋(𝑧) = ∑ 𝑥(𝑛) 𝑧 −𝑛
𝑛=−2
= 𝑥(−2)𝑧 −(−2) + 𝑥(−1)𝑧 −(−1) + 𝑥(0)𝑧 −(0) + 𝑥(1)𝑧 −(1) + 𝑥(2)𝑧 −(2)
= 1. 𝑧 2 + 3. 𝑧1 + 2.1 + (−3). 𝑧 −1 + 5. 𝑧 −2
∴ 𝑋(𝑧) = 𝑧 2 + 3𝑧 + 2 − 3𝑧 −1 + 5𝑧 −2

From the above equation we observe that 𝑋(𝑧) given finite value for any value of z except
𝑧 = 0 𝑎𝑛𝑑 𝑧 = ∞

So, for finite duration, non-causal signal, ROC is entire z plane except 𝒛 = 𝟎 𝒂𝒏𝒅 𝒛 = ∞

iv) Z-transform and ROC of a infinite duration and causal signal

Let a causal infinite duration signal 𝒙(𝒏) = 𝒂𝒏 𝒖(𝒏), 𝒘𝒉𝒆𝒓𝒆 𝒂 𝒊𝒔 𝒄𝒐𝒏𝒔𝒕𝒂𝒏𝒕

We know that z-transform of x(n) is,

∞

𝑋(𝑧) = ∑ 𝑥(𝑛) 𝑧 −𝑛
𝑛=−∞
𝒏
So for, 𝒙(𝒏) = 𝒂 𝒖(𝒏) ;
∞

𝑋(𝑧) = ∑ 𝑎𝑛 𝑢(𝑛) 𝑧 −𝑛
𝑛=−∞
1; 𝑛 ≥ 0
Now, 𝑢(𝑛) = {
0; 𝑛 < 0
−1 ∞
𝑛 −𝑛
∴ 𝑋(𝑧) = ∑ 𝑎 . 0. 𝑧 + ∑ 𝑎𝑛 . 1. 𝑧 −𝑛
𝑛=−∞ 𝑛=0
∞

= ∑ 𝑎𝑛 𝑧 −𝑛
𝑛=0
∞

= ∑(𝑎𝑧 −1 )𝑛
𝑛=0
Now, if |𝒂𝒛−𝟏 | < 1
|𝒂|
𝒐𝒓 <1
|𝒛|
𝒐𝒓|𝒂| < |𝒛|
 ∴ |𝒛| > |𝒂| then,
∞
1 𝑧
𝑋(𝑧) = ∑(𝑎𝑧 −1 )𝑛 = =
1 − 𝑎𝑧 −1 𝑧 − 𝑎
𝑛=0
−𝟏
And if |𝒂𝒛 | ≥ 𝟏 or|𝒛| < |𝒂|, then,
𝑋(𝑧) = ∑∞ −1 𝑛
𝑛=0(𝑎𝑧 ) = ∞
𝟏 𝒛
∴ 𝑿(𝒛) = 𝒁[𝒂𝒏 𝒖(𝒏)] = −𝟏
= , 𝒘𝒉𝒆𝒏 𝑹𝑶𝑪 𝒊𝒔 |𝒛| > |𝒂|
𝟏 − 𝒂𝒛 𝒛−𝒂

So, for infinite duration, causal signal, ROC is exterior of a circle with radius 𝒂 in z-
plane, i.e., |𝒛| > |𝒂|

v) Z-transform and ROC of a infinite duration and anti-causal signal

Let an anti-causal infinite duration signal 𝒙(𝒏) = 𝒃𝒏 𝒖(−𝒏 − 𝟏), 𝒘𝒉𝒆𝒓𝒆 𝒃 𝒊𝒔 𝒄𝒐𝒏𝒔𝒕𝒂𝒏𝒕

We know that z-transform of x(n) is,

∞

𝑋(𝑧) = ∑ 𝑥(𝑛) 𝑧 −𝑛
𝑛=−∞
So for, 𝒙(𝒏) = 𝒃𝒏 𝒖(−𝒏 − 𝟏) ;
∞

𝑋(𝑧) = ∑ 𝑏 𝑛 𝑢(−𝑛 − 1) 𝑧 −𝑛
𝑛=−∞
1; (−𝑛 − 1) ≥ 0 𝑜𝑟 − 𝑛 ≥ 1 𝑜𝑟 𝑛 ≤ −1
Now, 𝑢(−𝑛 − 1) = {
0; (−𝑛 − 1) < 0 𝑜𝑟 − 𝑛 < 1 𝑜𝑟 𝑛 > −1
−1 ∞

∴ 𝑋(𝑧) = ∑ 𝑏 𝑛 . 1. 𝑧 −𝑛 + ∑ 𝑏 𝑛 . 0. 𝑧 −𝑛
𝑛=−∞ 𝑛=0
−1

= ∑ 𝑏 𝑛 𝑧 −𝑛
𝑛=−∞
−1

= ∑ (𝑏𝑧 −1 )𝑛
𝑛=−∞

Put, −𝑛 = 𝑝, then 𝑛 = −𝑝 and for summation limit,

𝑤ℎ𝑒𝑛 𝑛 → −∞ 𝑡ℎ𝑒𝑛 𝑝 → ∞

𝑎𝑛𝑑 𝑤ℎ𝑒𝑛 𝑛 → −1 𝑡ℎ𝑒𝑛 𝑝 → 1

 1 ∞ ∞

∴ 𝑋(𝑧) = ∑ (𝑏𝑧 −1 )−𝑝 = ∑(𝑏 −1

𝑧) = [∑(𝑏 −1 𝑧)𝑝 ] − (𝑏 −1 𝑧)0
𝑝

𝑝=∞ 𝑝=1 𝑝=0

∞

= [∑(𝑏 −1 𝑧)𝑝 ] − 1
𝑝=0
−𝟏
Now, if |𝒃 𝒛| < 1
|𝒛|
𝒐𝒓 <1
|𝒃|
𝒐𝒓|𝒛| < |𝒃|
∴ 𝒘𝒉𝒆𝒏 |𝒛| < |𝒃| then,
∞
1 𝑏 𝑏−𝑏+𝑧 𝑧
𝑋(𝑧) = [∑(𝑏 −1 𝑧)𝑝 ] − 1 = −1
−1= −1= =
1−𝑏 𝑧 𝑏−𝑧 𝑏−𝑧 𝑏−𝑧
𝑝=0
And if |𝑏 −1
𝑧| ≥ 𝟏 or|𝒛| > |𝒃|, then,
∞
𝑋(𝑧) = ∑𝑛=0(𝑏 −1 𝑧)𝑛 = ∞
𝒛
∴ 𝑿(𝒛) = 𝒁[𝒃𝒏 𝒖(−𝒏 − 𝟏)] = , 𝒘𝒉𝒆𝒏 𝑹𝑶𝑪 𝒊𝒔 |𝒛| < |𝒃|
𝒃−𝒛
𝒛
𝒐𝒓, 𝒁[−𝒃𝒏 𝒖(−𝒏 − 𝟏)] = , 𝒘𝒉𝒆𝒏 𝑹𝑶𝑪 𝒊𝒔 |𝒛| < |𝒃|
𝒛−𝒃
So, for infinite duration, anti-causal signal, ROC is interior of a circle with radius 𝒃 in
z-plane, i.e., |𝒛| < |𝒃|

vi) Z-transform and ROC of a infinite duration and non-causal signal

Lets a non-causal infinite duration signal 𝒙(𝒏) = 𝒂𝒏 𝒖(𝒏) + 𝒃𝒏 𝒖(−𝒏 − 𝟏),

𝒘𝒉𝒆𝒓𝒆 𝒂 𝒂𝒏𝒅 𝒃 𝒊𝒔 𝒄𝒐𝒏𝒔𝒕𝒂𝒏𝒕

We know that z-transform of x(n) is,

∞

𝑋(𝑧) = ∑ 𝑥(𝑛) 𝑧 −𝑛
𝑛=−∞
So for, 𝒙(𝒏) = 𝒂𝒏 𝒖(𝒏) + 𝒃𝒏 𝒖(−𝒏 − 𝟏) ;
𝑋(𝑧) = 𝑍[𝑎𝑛 𝑢(𝑛) + 𝑏 𝑛 𝑢(−𝑛 − 1)] = 𝑍[𝑎𝑛 𝑢(𝑛)] + 𝑍[𝑏 𝑛 𝑢(−𝑛 − 1)]
As we know,
𝟏 𝒛
𝒁[𝒂𝒏 𝒖(𝒏)] = −𝟏
= , 𝒘𝒉𝒆𝒏 𝑹𝑶𝑪 𝒊𝒔 |𝒛| > |𝒂|
𝟏 − 𝒂𝒛 𝒛−𝒂
And,
𝒛
𝒁[𝒃𝒏 𝒖(−𝒏 − 𝟏)] = , 𝒘𝒉𝒆𝒏 𝑹𝑶𝑪 𝒊𝒔 |𝒛| < |𝒃|
𝒃−𝒛
 ∴ 𝒁[𝒂𝒏 𝒖(𝒏) + 𝒃𝒏 𝒖(−𝒏 − 𝟏)] = 𝑍[𝑎𝑛 𝑢(𝑛)] + 𝑍[𝑏 𝑛 𝑢(−𝑛 − 1)]
𝑧 𝑧
= ; 𝑤ℎ𝑒𝑛 𝑅𝑂𝐶 ∶ |𝑧| > |𝑎| + , 𝑤ℎ𝑒𝑛 𝑅𝑂𝐶: |𝑧| < |𝑏|
𝑧−𝑎 𝑏−𝑧
Now,
• Case I: if 𝒂 > 𝑏 and |𝑧| > |𝑎| 𝑎𝑛𝑑 |𝑧| < |𝑏|
Then there has no common region or connected region of ROC in z-plane,
So when 𝒂 > 𝑏, ROC does not exist and 𝑍[𝑎𝑛 𝑢(𝑛) + 𝑏 𝑛 𝑢(−𝑛 − 1)] cannot be find out.

• Case II: if 𝒂 < 𝑏 and |𝑧| > |𝑎| 𝑎𝑛𝑑 |𝑧| < |𝑏|
Then there has a common region or connected region of ROC in z-plane, and the
resultant ROC of the function 𝑿(𝒛) will be |𝒂| < |𝒛| < |𝒃|

𝒛 𝒛
∴ 𝒁[𝒂𝒏 𝒖(𝒏) + 𝒃𝒏 𝒖(−𝒏 − 𝟏)] = + , 𝒘𝒉𝒆𝒏 𝑹𝑶𝑪 𝒊𝒔 |𝒂| < |𝒛| < |𝒃| 𝒂𝒏𝒅 𝒂 < 𝑏
𝒛−𝒂 𝒃−𝒛

So, for infinite duration, non-causal signal, ROC is the intermediate region between two
circle with radius 𝒂 and 𝒃 where 𝒂 < 𝑏 , i.e., ROC: |𝒂| < |𝒛| < |𝒃|.