Chapter 4

Continuous-Time Fourier Transform

1
 Introduction
• Periodic signals are represented as linear combination of harmonically related complex exponentials (Fourier Series).
• Non-periodic (Aperiodic) signals are represented by complex exponentials using Fourier Transform.

𝑖𝑓 𝑤𝑒 𝑙𝑒𝑡 𝑇 → ∞ ⇒ 𝜔0 → 0 𝑐𝑜𝑛𝑡𝑖𝑛𝑢𝑜𝑢𝑠 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 𝜔0 = 𝑑𝜔 ෍ becomes න

𝑥(𝑡) 𝑥𝑇 (𝑡)
periodic extension

𝑇 −𝑇 𝑇 2𝑇
Non-periodic signal periodic signal formed by periodic extension of 𝑥(𝑡)
Graphical interpretation

ℱ𝑇

Fourier Transform of 𝑥(𝑡) Inverse Fourier Transform of 𝑋(𝑗𝜔)

X(𝜔) is the frequency domain specification of 𝑥(𝑡) (Spectrum) X(𝜔) magnitude spectrum
FT
The Fourier transform pair 𝑥(𝑡) X(𝜔) X(𝜔) is complex, and it can be expressed as: X 𝜔 = X(𝜔) 𝑒 𝑗∅ 𝜔 ∅ 𝜔 phase spectrum
2
 Fourier Transform - Example 1
𝛿 𝑡
1
Consider the unit impulse function 𝑥 𝑡 = 𝛿(𝑡)
1 𝑡=0
𝛿 𝑡 =ቊ 𝑡
Find its Fourier transform. 0 𝑡≠0 0

Answer

Fourier transform

+∞
ℱ 𝛿 𝑡 = න 𝛿 𝑡 𝑒 −𝑗𝜔𝑡 𝑑𝑡 = 𝛿 0 𝑒 −𝑗𝜔 0 = 1 𝑓𝑜𝑟 𝑎𝑙𝑙 𝜔
−∞

𝑋 𝜔
1

0 𝜔

3
 Fourier Transform - Example 2
𝑒 −𝑎𝑡 𝑡≥0
Consider the exponential signal 𝑥 𝑡 =𝑒 −𝑎𝑡
𝑢 𝑡 =ቊ ,𝑎 > 0
0 𝑡<0
Find its Fourier transform.
Answer

Fourier transform

1 𝑎 − 𝑗𝜔 𝑎 − 𝑗𝜔 𝑎 𝜔
𝑋 𝜔 = × = 2 = − 𝑗
𝑎 + 𝑗𝜔 𝑎 − 𝑗𝜔 𝑎 + 𝜔 2 𝑎2 + 𝜔 2 𝑎2 + 𝜔 2

Module
sin 𝜃
tan 𝜃 =
cos 𝜃
𝜋
tan =∞
Phase 2
𝜋
tan − = −∞
2 4
 Fourier Transform - Example 3
Find Fourier transform of: Time-domain

𝑥 𝑡 = 𝑒 −𝑎 𝑡 , 𝑎 > 0 𝑒 −𝑎 𝑡 𝑒 −𝑎𝑡
= ቊ +𝑎𝑡 𝑡≥0
𝑒 𝑡<0
Answer
𝑡 𝑡≥0
The Fourier transform of the signal is: 𝑡 =ቊ
−𝑡 𝑡<0

Frequency-domain

5
 Fourier Transform - Example 4
Find Fourier transform of: 1
1, 𝑡 < 𝑇1
𝑥 𝑡 =ቊ
0, 𝑡 > 𝑇1
Answer
2𝑗 𝑠𝑖𝑛 𝑥 = 𝑒 𝑗𝑥 − 𝑒 −𝑗𝑥

sin 𝑥
𝑠𝑖𝑛𝑐 𝑥 =
𝑥

sin 𝜔𝑇1
2𝑗 𝑒 𝑗𝜔𝑇1 − 𝑒 −𝑗𝜔𝑇1 2𝑇1 = 2𝑇1 𝑠𝑖𝑛𝑐 𝜔𝑇1
𝜔𝑇1
𝑗𝜔 2𝑗
6
 Fourier Transform - Example 4
Find the inverse Fourier transform of:
1, 𝜔 < 𝑊
𝑋 𝑗𝜔 = ቊ
0, 𝜔 > 𝑊
the inverse Fourier transform
1 ∞
𝑥 𝑡 = න 𝑋 𝑗𝜔 𝑒 𝑗𝜔𝑡 𝑑𝜔
2𝜋 −∞
𝑊 𝑗𝜔𝑡 𝑊
1 𝑗𝜔𝑡
1 𝑒
𝑥 𝑡 = න 1 𝑒 𝑑𝜔 =
2𝜋 −𝑊 2𝜋 𝑗𝑡 −𝑊
1 𝑒 𝑗𝑊𝑡 𝑒 −𝑗𝑊𝑡 2 𝑒 𝑗𝑊𝑡
− 𝑒 −𝑗𝑊𝑡 𝑊 𝑠𝑖𝑛 𝑊𝑡
𝑥 𝑡 = − = =
2𝜋 𝑗𝑡 𝑗𝑡 2𝜋𝑡 𝑗2 𝜋 𝑊𝑡 7
 Problem 1
Use the Fourier transform analysis equation to calculate the Fourier transforms of :

Answer 𝜏 =𝑡−1
𝑡 =𝜏+1
(a) 𝑑𝜏 = 𝑑𝑡 (b)

𝑒 −𝑗𝜔(𝜏+1) = 𝑒 −𝑗𝜔 𝑒 −𝑗𝜔𝜏

𝑦
𝑗 𝑡𝑎𝑛−1 𝑥
𝑧 = 𝑥 + 𝑗𝑦 = 𝑧 𝑒
1 1
𝑧1 𝑧1 𝑒 𝑗𝜃1 𝑧1 𝑗 𝜃1 −𝜃2 𝑒 −𝑗𝜔 +
= = 𝑒 (2 − 𝑗𝜔) (2 + 𝑗𝜔)
2 + 𝑗𝜔 = 4+ 𝜔2𝑒
𝑗𝑡𝑎𝑛 −1
𝜔
2 𝑧2 𝑧2 𝑒 𝑗𝜃2 𝑧2 8
 Problem 2
Use the Fourier transform analysis equation to calculate the Fourier transforms of :
𝑥 𝑡 =
Answer

Some useful Fourier transform:

9
 Fourier Transform for Periodic Signals
Obtain the Fourier transform of a periodic signal 𝑥 𝑡 directly from its Fourier series 𝑎𝑘
Inverse FT
IF Than
Proof ∞ ∞ ∞
1 ∞ 2𝜋 ∞
𝑥 𝑡 = න ෍ 𝑎𝑘 2𝜋𝛿(𝜔 − 𝑘𝜔0 ) 𝑒 𝑗𝜔𝑡 𝑑𝜔 = ෍ 𝑎𝑘 න 𝛿(𝜔 − 𝑘𝜔0 )𝑒 𝑗𝜔𝑡 𝑑𝜔 = ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0𝑡
2𝜋 −∞ 2𝜋 −∞
𝑘=−∞ 𝑘=−∞ 𝑘=−∞

Example 5
Find Fourier transform of x(t)

The Fourier series coefficients of the above periodic signal:

See slide (6)

Fourier transform Fourier transform of a symmetric
periodic square wave

10
 Example 4
Obtain the Fourier transform of
Solution:

Fourier series coefficients:

Fourier transform

11
 Problem 5
𝜋
Determine the Fourier transform of the periodic signal. 𝑥 𝑡 = 1 + cos 6𝜋𝑡 +
8
Solution:

1 𝑗𝜋 𝑗 1 −𝑗𝜋 𝑗(−1)6𝜋𝑡
𝑥 𝑡 =1 𝑒 𝑗(0)6𝜋𝑡 + 𝑒 8 𝑒 1 6𝜋𝑡 + 𝑒 8 𝑒
𝑥 𝑡 2 2
𝑎𝑘 = 0
𝑎0 𝑎1 𝑎−1 all other k
∞
The Fourier transform of the periodic signal 𝑋 𝑗𝜔 = ෍ 𝑎𝑘 2𝜋𝛿 𝜔 − 𝑘𝜔0
𝑘=−∞
𝜋 𝜋
𝑗 −𝑗
= 𝑎0 2𝜋𝛿 𝜔 − (0)𝜔0 + 𝑎1 2𝜋𝑒 8𝛿 𝜔 − (1)𝜔0 + 𝑎−1 2𝜋𝑒 8𝛿 𝜔 − (−1)𝜔0
𝜋 𝜋
𝑗 −𝑗
= 2𝜋𝛿 𝜔 + 𝜋𝑒 8𝛿 𝜔 − 6𝜋 + 𝜋𝑒 8𝛿 𝜔 + 6𝜋 12
 Problem 6
Use the Fourier synthesis equations to determine the Inverse Fourier transform of

Solution:

𝑋1 𝑗𝜔
𝜋 2𝜋 𝜋

−4𝜋 0 4𝜋 𝜔

𝑥 𝑡

13
 Problem 7
Determine the Fourier transform of the a DC Value (Constant Function) 𝑥 𝑡 = 𝐴0
Solution:
∞ ∞ ∞
𝑥 𝑡 = 𝐴0 is not an absolutely integrable signal න 𝑥(𝑡) 𝑑𝑡 < ∞ 𝑤𝑒 ℎ𝑎𝑣𝑒 න 𝐴0 𝑑𝑡 = 𝐴0 𝑡ቚ =∞
−∞ −∞ −∞

1 ∞ 𝑗𝜔𝑡
1 ∞
𝑖𝑓 𝑤𝑒 𝑡𝑎𝑘𝑒 𝑋 𝜔 = 𝐴0 𝛿 𝜔 and compute the inverse 𝑥 𝑡 = න 𝑋(𝜔) 𝑒 𝑑𝜔 = න 𝐴0 𝛿 𝜔 𝑒 𝑗𝜔𝑡 𝑑𝜔
2𝜋 −∞ 2𝜋 −∞

𝐴0 ∞ 𝑗𝜔𝑡
𝐴0 𝑗0𝑡
𝐴0 1 𝜔=0
𝑥 𝑡 = න 𝛿 𝜔 𝑒 𝑑𝜔 = 𝛿 0 𝑒 = With 𝛿 𝜔 = ቊ
2𝜋 −∞ 2𝜋 2𝜋 0 𝜔≠0

𝐼ℱ ℱ
𝐴0 𝐴0
𝐴0 𝛿 𝜔 2𝜋 𝐴0 𝛿 𝜔
2𝜋

14
 Problem 8
1 𝑡≥0
Determine the Fourier transform of the Unit Step Function 𝑢 𝑡 = ቊ
0 𝑡<0
Solution:
𝑥 𝑡 = 𝑢 𝑡 is not an absolutely integrable signal
∞ ∞ ∞
න 𝑥(𝑡) 𝑑𝑡 < ∞ 𝑤𝑒 ℎ𝑎𝑣𝑒 න 𝑢(𝑡) 𝑑𝑡 = 𝑡ቚ = ∞
−∞ −∞ 0

ℱ
Even and odd parts of 𝑢 𝑡 : 𝑢 𝑡 = 𝑢𝑒 𝑡 + 𝑢𝑜 𝑡 𝑋𝑒 𝜔 + 𝑋𝑜 𝜔

𝑢 𝑡 + 𝑢 −𝑡 ℱ
Even 𝑢𝑒 𝑡 = = 1ൗ2 𝑋𝑒 𝜔 = 2𝜋𝐴0 𝛿 𝜔 = 𝜋𝛿 𝜔
2
ℱ 1 1
odd 2 𝑢𝑜 𝑡 = 𝑢 𝑡 − 𝑢 −𝑡 = lim 𝑒 −𝑎𝑡 𝑢 𝑡 − lim 𝑒 𝑎𝑡 𝑢 −𝑡 2𝑋𝑜 𝜔 = lim −
𝑎→0 𝑎→0 𝑎→0 𝑎 + 𝑗𝜔 𝑎 − 𝑗𝜔

−2𝑗𝜔 −2𝑗𝜔 1 1
2𝑋𝑜 𝜔 = lim 2 = =2 𝑋𝑜 𝜔 =
𝑎→0 𝑎 + 𝜔 2 𝜔2 𝑗𝜔 𝑗𝜔

ℱ 1
𝑋 𝜔 = 𝑋𝑒 𝜔 + 𝑋𝑜 𝜔 𝑢 𝑡 𝜋𝛿 𝜔 +
𝑗𝜔 15
 Properties of CT Fourier Transform
• Give insight of the relationship between the time-domain and frequency-domain descriptions of a signal.

• Are useful in reducing the complexity of the evaluation of Fourier transforms or inverse transforms.

1. Linearity ℱ ℱ
𝑥 𝑡 𝑋 𝑗𝜔
Then 𝑎 𝑥 𝑡 + 𝑏 𝑦 𝑡 𝑎 𝑋 𝑗𝜔 + 𝑏 𝑌 𝑗𝜔
If ℱ
𝑦 𝑡 𝑌 𝑗𝜔
Proof ℱ
𝑥 𝑡 ℱ 𝑥 𝑡 = 𝑋(𝜔)
2. Time shifting 1 ∞ 1 ∞
𝑗𝜔𝑡
𝑥 𝑡 = න ℱ 𝑥 𝑡 𝑒 𝑑𝜔 = න 𝑋(𝜔)𝑒 𝑗𝜔𝑡 𝑑𝜔
2𝜋 −∞ 2𝜋 −∞
ℱ 1 ∞
𝑥 𝑡 𝑋 𝑗𝜔 Proof 𝑥 𝑡 − 𝑡0 = න 𝑋(𝜔)𝑒 𝑗𝜔(𝑡−𝑡0) 𝑑𝜔
2𝜋 −∞
ℱ 1 ∞
𝑥 𝑡 − 𝑡0 𝑒 −𝑗𝜔𝑡0 𝑋 𝑗𝜔 = න 𝑋(𝜔)𝑒 −𝑗𝜔𝑡0 𝑒 𝑗𝜔𝑡 𝑑𝜔
2𝜋 −∞
If a signal is time-shifted, the magnitude of the 1 ∞
Fourier transform does not change; only there = න ℱ 𝑥 𝑡 − 𝑡0 𝑒 𝑗𝜔𝑡 𝑑𝜔
2𝜋 −∞
is a phase-shift in the Fourier transform.
ℱ
𝑥 𝑡 − 𝑡0 ℱ 𝑥 𝑡 − 𝑡0 = 𝑋(𝜔)𝑒 −𝑗𝜔𝑡0 16
 Example 51
Evaluate the Fourier transform of 𝑥(𝑡)
Solution:
𝑥0 𝑡 = 𝑥(𝑡 + 2.5) 𝑥 𝑡 = 𝑥0 (𝑡 − 2.5)

𝑆ℎ𝑖𝑓𝑡𝑖𝑛𝑔 𝑏𝑦 𝑡0 = 2.5

2 𝑠𝑖𝑛 𝜔ൗ2 ℱ 2 𝑠𝑖𝑛 3 𝜔ൗ2

By Decomposing ℱ 𝑋1 𝑗𝜔 = 𝑋2 𝑗𝜔 =
𝜔 𝜔
into simpler signals

17
 Example 52
From Example 3 and the signals of the previous slide, we get:

2 𝑠𝑖𝑛 𝜔ൗ2 2 𝑠𝑖𝑛 3 𝜔ൗ2

𝑋1 𝑗𝜔 = and 𝑋2 𝑗𝜔 =
𝜔 𝜔

By linearity and time shifting properties:

1
𝑋 𝑗𝜔 = ℱ 𝑥1 𝑡 − 2.5 + ℱ 𝑥2 𝑡 − 2.5
2
1 −2.5𝑗𝜔
= 𝑒 ℱ 𝑥1 𝑡 + 𝑒 −2.5𝑗𝜔 ℱ 𝑥2 𝑡
2
1 −2.5𝑗𝜔 2 𝑠𝑖𝑛 𝜔ൗ2 2 𝑠𝑖𝑛 3 𝜔ൗ
2
= 𝑒 + 𝑒 −2.5𝑗𝜔
2 𝜔 𝜔
𝑠𝑖𝑛 𝜔ൗ + 2 𝑠𝑖𝑛 3 𝜔ൗ
= 𝑒 −2.5𝑗𝜔 2 2
𝜔
18
 Properties of CT Fourier Transform
3. Conjugate Symmetry Property
ℱ ℱ
If the signal 𝑥 𝑡 is complex then: If 𝑥 𝑡 𝑋 𝜔 Then 𝑥∗ 𝑡 𝑋 ∗ −𝜔
Proof
∞ ∗ ∞ Replace 𝜔 by -𝜔
∞
𝑋∗ 𝜔 = 𝑋 𝜔 ∗= න 𝑥(𝑡)𝑒 −𝑗𝜔𝑡 𝑑𝑡 =න 𝑥 ∗ (𝑡)𝑒 𝑗𝜔𝑡 𝑑𝑡 𝑋 −𝜔 = න 𝑥 ∗ (𝑡)𝑒 −𝑗𝜔𝑡 𝑑𝑡 = ℱ 𝑥 ∗ (𝑡)
∗
−∞ −∞ −∞

If the signal 𝑥 𝑡 is real then: 𝑋 𝜔 = 𝑋∗

−𝜔
𝑧∗ = 𝑧
𝑋 𝜔 = 𝑋 −𝜔 Magnitude Spectrum is even
𝑥 𝑡 = 𝑥∗ 𝑡 𝑋 𝜔 = 𝑋∗ −𝜔 ቊ ቊ
∠𝑋 𝜔 = ∠𝑋 ∗ −𝜔 ∠𝑧 ∗ = −∠𝑧 ∠𝑋 𝜔 = −∠𝑋 −𝜔 Phase Spectrum is odd
From positive frequencies we can determine magnitude and phase of 𝑋 𝑗𝜔 for negative frequencies

For 𝑥 𝑡 real and even: 𝑋 𝜔 = 𝑅 𝜔 + 𝑗𝐼 𝜔

𝑥∗ 𝑡 =𝑥 𝑡 Real
𝑋∗ −𝜔 = 𝑋 𝜔 𝑋 −𝜔 + 𝑋 ∗ −𝜔 = 2𝑋 𝜔 𝑋 ∗ −𝜔 = 𝑅 −𝜔 − 𝑗𝐼 −𝜔

Even 2𝑅 −𝜔 = 2𝑅 𝜔 + 2𝑗𝐼 𝜔 𝐼 𝜔 =0 𝑋 −𝜔 = 𝑅 −𝜔 + 𝑗𝐼 −𝜔
𝑥 𝑡 = 𝑥 −𝑡 𝑋 −𝜔 = 𝑋 𝜔
𝑋 𝜔 =𝑅 𝜔 =𝑅 −𝜔 is purely Real
For 𝑥 𝑡 real and odd: and even For 𝑥 𝑡 = 𝑥𝑒 𝑡 + 𝑥𝑜 𝑡
𝑋 𝑗𝜔 is purely imaginary and odd 𝑋 ∗ −𝜔 − 𝑋 −𝜔 = 2𝑋 𝜔 ℱ
𝑥∗ 𝑡 = 𝑥 𝑡
Real
𝑋 ∗ −𝜔 = 𝑋 𝜔 −2𝑗𝐼 −𝜔 = 2𝑅 𝜔 + 2𝑗𝐼 𝜔 𝑅 𝜔 =0 Even 𝑥 𝑡 Re 𝑋 𝜔
𝑋 𝜔 =𝑗𝐼 𝜔 =−𝑗𝐼 −𝜔 is purely ℱ
𝑥 𝑡 = −𝑥 −𝑡
odd
𝑋 −𝜔 = −𝑋 𝜔 Odd 𝑥 𝑡 j Im 𝑋 𝜔
imaginary and odd 19
 Example 6
Evaluate the Fourier transform of 𝑥 𝑡 = 𝑒 −𝑎 𝑡 for 𝑎 > 0

From Example 6: 𝑥 𝑡 = 𝑒 −𝑎 𝑡 for 𝑎 > 0

𝑒 −𝑎𝑡 𝑢 𝑡 + 𝑒 𝑎𝑡 𝑢 −𝑡
𝑥 𝑡 = 𝑒 −𝑎 𝑡 = 𝑒 −𝑎𝑡 𝑢 𝑡 + 𝑒 𝑎𝑡 𝑢 −𝑡 = 2 = 2 𝐸𝑣𝑒𝑛 𝑒 −𝑎𝑡 𝑢 𝑡
2
𝑥1 𝑡 + 𝑥1 −𝑡
𝐸𝑣𝑒𝑛 𝑥1 𝑡 =
ℱ 1 2
we have 𝑥1 𝑡 = 𝑒 −𝑎𝑡 𝑢(𝑡) 𝑋 𝑗𝜔 =
𝑎 + 𝑗𝜔

𝑒 −𝑎𝑡 𝑢 𝑡 is real; from symmetric property,

ℱ 1 𝑎 − 𝑗𝜔
2 𝐸𝑣𝑒𝑛 𝑒 −𝑎𝑡 𝑢 𝑡 2 𝑅𝑒 = 2𝑅𝑒 2
𝑎 + 𝑗𝜔 𝑎 + 𝜔2

2𝑎 mirror image
𝑋 𝑗𝜔 = 2
𝑎 + 𝜔2 20
 Properties of CT Fourier Transform
Proof
1 ∞
4. Differentiation property 𝑥 𝑡 = න 𝑋 𝑗𝜔 𝑒 𝑗𝜔𝑡 𝑑𝜔
2𝜋 −∞
ℱ 𝑑𝑥(𝑡) 1 ∞ 𝑑 𝑗𝜔𝑡
If 𝑥 𝑡 𝑋 𝑗𝜔 = න 𝑋(𝑗𝜔) 𝑒 𝑑𝜔
𝑑𝑡 2𝜋 −∞ 𝑑𝑡

𝑑𝑥(𝑡) 1 ∞
𝑑𝑥(𝑡) ℱ 𝑗𝜔 𝑋 𝑗𝜔 𝑑𝑡
= න 𝑋(𝑗𝜔)𝑗𝜔 𝑒 𝑗𝜔𝑡 𝑑𝜔
2𝜋 −∞
Then 𝑑𝑡
𝑑𝑥(𝑡) 1 ∞
= න 𝑗𝜔 𝑋(𝑗𝜔) 𝑒 𝑗𝜔𝑡 𝑑𝜔
𝑑𝑡 2𝜋 −∞

5. Integration property
Proof 𝑡 ∞
𝑥 𝑡 = න 𝑔(𝜏) 𝑑𝜏 = න 𝑔 𝜏 𝑢(𝑡 − 𝜏)𝑑𝜏
−∞ −∞ 𝑖𝑠 1 𝑓𝑜𝑟 𝜏 ≤ 𝑡
ℱ 𝑥 𝑡 =𝑔 𝑡 ∗𝑢 𝑡 ↔𝑋 𝜔 =𝐺 𝜔 𝑈 𝜔
If 𝑥 𝑡 𝑋 𝑗𝜔 1
DC or average value ℱ 𝑢(𝑡) = + 𝜋𝛿 𝜔 𝑠𝑒𝑒 𝑝𝑟𝑜𝑏𝑙𝑒𝑚 7
𝑗𝜔
𝑡 ℱ 1 𝑋 𝜔 =𝐺 𝜔
1
+ 𝜋𝛿 𝜔 𝑋 𝜔 =
𝐺 𝜔
𝑗𝜔
+ 𝜋𝐺 0 𝛿 𝜔
Then න 𝑥(𝜏) 𝑑𝜏 𝑋 𝑗𝜔 + 𝜋𝑋 0 𝛿(𝜔) 𝑗𝜔
−∞ 𝑗𝜔 21
 Example 7
Determine the Fourier transform of the unit step function.

ℱ
For unit impulse 𝐺 𝑗𝜔 = 1

𝑡 ℱ 𝐺 𝑗𝜔
Now, 𝑥 𝑡 = 𝑢(𝑡) = න 𝛿(𝜏) 𝑑𝜏 + 𝜋𝐺 0 𝛿(𝜔)
−∞ 𝑗𝜔
1 1
𝑋 𝑗𝜔 = + 𝜋𝛿(𝜔) 𝐺 𝑗𝜔 = 1 → 𝐺 0 = 1
𝑗𝜔
Also, we observe that
0
𝑑𝑢(𝑡) ℱ 1
𝛿 𝑡 = 𝑗𝜔 + 𝜋𝛿(𝜔) = 1 + 𝜋 𝑗𝜔 𝛿 𝜔 = 1 𝜔𝛿 𝜔 =0=ቊ
0 ∙1 𝜔=0
𝑑𝑡 𝑗𝜔 𝜔∙0
22
𝜔≠0
 Properties of CT Fourier Transform
6.Time and Frequency Scaling

ℱ ℱ 1 𝑗𝜔
If 𝑥(𝑡) 𝑋 𝑗𝜔 Than 𝑥(𝑎 𝑡) 𝑋
𝑎 𝑎
Proof
1
𝜏 = 𝑎𝑡 ⇒ 𝑑𝜏 = 𝑎 𝑑𝑡 ⇒ 𝑑𝑡 = 𝑑𝜏
∞
𝑎 ∞ 𝜏
−𝑗𝜔𝑎 1 1 ∞ 𝜔
−𝑗 𝑎 𝜏
ℱ 𝑥 𝑎𝑡 = න 𝑥 𝑎𝑡 𝑒 −𝑗𝜔𝑡 𝑑𝑡 =න 𝑥 𝜏 𝑒 𝑑𝜏 = න 𝑥 𝜏 𝑒 𝑑𝜏
−∞ −∞ 𝑎 𝑎 −∞
1 ∞ 𝜔
−𝑗 𝑎 𝜏
𝑓𝑜𝑟 𝑎 > 0, න 𝑥 𝜏 𝑒 𝑑𝜏
𝑎 −∞
=
1 −∞ 𝜔
−𝑗 𝑎 𝜏 1 ∞ 𝜔
−𝑗 𝑎 𝜏
𝑓𝑜𝑟𝑎 < 0, න 𝑥 𝜏 𝑒 𝑑𝜏 = − න 𝑥 𝜏 𝑒 𝑑𝜏 , For a < 0: 𝑡 = −∞ ⇒ 𝜏 = ∞; 𝑎𝑡 𝑡 = ∞ ⇒ 𝜏 = −∞
𝑎 ∞ 𝑎 −∞

ℱ Reversing a signal in time reverses

In particular, 𝑥(−𝑡) 𝑋 −𝑗𝜔 its Fourier transform also.
23
 Property 7: Duality

• The FT and IFT relations are ℱ

similar

• This symmetry leads to

duality property of the
Fourier transform.
ℱ

To the Fourier transform pair Corresponds the following dual Fourier transform pair:

If 𝑥(𝑡) 𝐺 𝜔 Than 𝐺(𝑡) 2𝜋 𝑥 −𝑗𝜔

24
 Example 8
Use the duality property to determine 𝐺(𝑗𝜔), the Fourier transform of 𝑔 𝑡 = 2/(1 + 𝑡 2 ).

From example 2: ℱ 2𝑎
𝑥 𝑡 = 𝑒 −𝑎 𝑡 ,𝑎 > 0 𝑋 𝑗𝜔 = 2
𝑎 + 𝜔2

For a = 1, 𝑥 𝑡 = 𝑒 −𝑡 ℱ 2
𝑋 𝑗𝜔 =
1 + 𝜔2
The synthesis equation for this FT pair is:
Multiplying by 2𝜋
∞
1 ∞ 2 and replacing t by -t
2
𝑒− 𝑡 = න 𝑒 𝑗𝜔𝑡 𝑑𝜔 2𝜋 𝑒− 𝑡 =න 𝑒 −𝑗𝜔𝑡 𝑑𝜔
2𝜋 −∞ 1 + 𝜔 2 −∞ 1 + 𝜔2
interchanging 𝑔 𝑡
We obtain FT
𝑡 and 𝜔
∞
2 analysis equation 2 ℱ
2𝜋 𝑒 − 𝜔 = න 𝑒 −𝑗𝜔𝑡 𝑑𝑡
(1 + 𝑡 2 ) 2𝜋 𝑒 − 𝜔
−∞ 1 + 𝑡2
𝐺(𝑗𝜔)
𝐺 𝑗𝜔 = 2𝜋 𝑒 − 𝜔 25
 Property 8: Convolution

ℱ
For an LTI system: 𝑦 𝑡 = ℎ(𝑡) ∗ 𝑥(𝑡) 𝑌 𝑗𝜔 = 𝑋 𝑗𝜔 𝐻 𝑗𝜔

A convolution in time domain implies a multiplication in Fourier domain.

Proof
ℱ ∞ ∞
𝑦 𝑡 = ℎ(𝑡) ∗ 𝑥(𝑡) 𝑌 𝜔 =න 𝑦(𝑡) 𝑒 −𝑗𝜔𝑡 𝑑𝑡 =න ℎ(𝑡) ∗ 𝑥(𝑡) 𝑒 −𝑗𝜔𝑡 𝑑𝑡
−∞ −∞
∞ ∞ ∞ ∞
𝑌 𝜔 =න න 𝑥 𝜏 ℎ 𝑡 − 𝜏 𝑑𝜏 𝑒 −𝑗𝜔𝑡 𝑑𝑡 = න 𝑥 𝜏 න ℎ 𝑡 − 𝜏 𝑒 −𝑗𝜔𝑡 𝑑𝑡 𝑑𝜏
−∞ −∞ −∞ −∞

∞
We have න ℎ 𝑡 − 𝜏 𝑒 −𝑗𝜔𝑡 𝑑𝑡 = ℱ ℎ 𝑡 − 𝜏 = 𝑒 −𝑗𝜔𝜏 𝐻 𝜔 From the time-shifting property
−∞

∞ ∞
𝑌 𝜔 = න 𝑥 𝜏 𝑒 −𝑗𝜔𝜏 𝐻 𝜔 𝑑𝜏 = 𝐻 𝜔 න 𝑥 𝜏 𝑒 −𝑗𝜔𝜏 𝑑𝜏 = 𝐻 𝜔 𝑋 𝜔
−∞ −∞

ℱ 𝑌 𝜔 =𝐻 𝜔 𝑋 𝜔
𝑦 𝑡 = ℎ(𝑡) ∗ 𝑥(𝑡)
26
 Property 9: Differentiation in Frequency Domain
We have the differentiation in frequency domain property 𝑑
𝑡𝑥 𝑡 ℱ 𝑗 𝑋 𝑗𝜔
𝑋 𝑗𝜔
ℱ 𝑥 𝑡 𝑗
𝑑
𝑋 𝑗𝜔
ℱ 𝑡𝑥 𝑡 𝑑𝜔
𝑑𝜔

Proof: We differentiate the Fourier transform of 𝑥 𝑡 with respect to 𝜔 to get:

𝑑
𝑋 𝑗𝜔 =
𝑑 ∞ ∞
න 𝑥 𝑡 𝑒 −𝑗𝜔𝑡 𝑑𝑡 = න 𝑥 𝑡
𝑑 ∞
𝑒 −𝑗𝜔𝑡 𝑑𝑡 = න 𝑥 𝑡 −𝑗𝑡 𝑒 −𝑗𝜔𝑡 𝑑𝑡 𝑡𝑥 𝑡
ℱ 𝑗
𝑑
𝑋 𝑗𝜔
𝑑𝜔 𝑑𝜔 −∞ −∞ 𝑑𝜔 −∞ 𝑑𝜔

Example: Using frequency derivative property of Fourier transform, find the Fourier transform of function 𝑡𝑒 −2𝑡 𝑢 𝑡

ℱ𝑇 1 ℱ𝑇 𝑑
Let 𝑥 𝑡 = 𝑒 −2𝑡 𝑢 𝑡 𝑋 𝜔 = By using the frequency derivative property, We have 𝑡∙𝑥 𝑡 𝑗 𝑋 𝜔
2 + 𝑗𝜔 𝑑𝜔

ℱ𝑇 𝑑 1 0 2 + 𝑗𝜔 − (𝑗)(1) 1
𝑡𝑒 −2𝑡 𝑢 𝑡 𝑗 =𝑗 =
𝑑𝜔 2 + 𝑗𝜔 2 + 𝑗𝜔 2 2 + 𝑗𝜔 2

ℱ𝑇 1
𝑡𝑒 −2𝑡 𝑢 𝑡 2
2 + 𝑗𝜔

27
 Find 𝐻 𝑗𝜔 for the LTI systems
Example 9: Differentiator
ℱ
From differential property,

Frequency response
Y 𝑗𝜔 = 𝐻 𝑗𝜔 𝑋 𝑗𝜔 𝐻 𝑗𝜔 = 𝑗𝜔 of a differentiator.
Example 10: Integrator
𝑡
ℎ 𝑡 = න 𝛿 𝑡 𝑑𝑡 = 𝑢 𝑡
−∞
The impulse response of this system is a unit step, u(t).
ℱ

28
 Example 91
Find the response, 𝑦 𝑡 of an LTI system, if
ℱ
𝑦 𝑡 = 𝑥 𝑡 ∗ ℎ(𝑡) 𝑌 𝑗𝜔 = 𝑋 𝑗𝜔 𝐻(𝑗𝜔)
ℱ 1 ℱ 1
𝑥 𝑡 = 𝑒 −𝑏𝑡 𝑢 𝑡 𝑋 𝑗𝜔 = and ℎ 𝑡 = 𝑒 −𝑎𝑡 𝑢 𝑡 𝐻 𝑗𝜔 =
𝑎 + 𝑗𝜔
𝑏 + 𝑗𝜔

𝐼ℱ 1
𝑒 −𝑎𝑡 − 𝑒 −𝑏𝑡 𝑢 𝑡 For 𝑏 ≠ 𝑎
𝑏−𝑎
1 1 1
If b = a, the partial fraction expansion is not valid 𝑌 𝑗𝜔 = = 2
𝑎 + 𝑗𝜔 𝑎 + 𝑗𝜔 𝑎 + 𝑗𝜔

We know:

1
ℱ 𝑦 𝑡 = ൞𝑏 − 𝑎
𝑒 −𝑎𝑡 − 𝑒 −𝑏𝑡 𝑢 𝑡 for 𝑏 ≠ 𝑎
𝑡𝑒 −𝑎𝑡 𝑢 𝑡 for 𝑏 = 𝑎 29
Some Important CT Relationship

30
 Differential Equation and Frequency Response of a System
differential equation

The system frequency response 𝐻 𝑗𝜔

Convolution
property, frequency response
𝑌 𝑗𝜔
𝑦 𝑡 =ℎ 𝑡 ∗𝑥 𝑡 𝑌 𝑗𝜔 = 𝐻 𝑗𝜔 𝑋 𝑗𝜔 𝐻 𝑗𝜔 =
𝑋 𝑗𝜔
Example 10
𝑥 𝑡 ℱ 𝑋 𝑗𝜔
Consider the system characterized 𝑑𝑥 𝑡
by the differential equation 𝑑𝑡
ℱ 𝑗𝜔 𝑋 𝑗𝜔

the frequency response :

ℱ 𝑗𝜔 𝑌 𝑗𝜔 + 𝑎 𝑌 𝑗𝜔 = 𝑋 𝑗𝜔
impulse response
𝑌 𝑗𝜔 1
(𝑗𝜔 + 𝑎) 𝑌 𝑗𝜔 = 𝑋 𝑗𝜔 𝐻 𝑗𝜔 = = ℎ 𝑡 = 𝑒 −𝑎𝑡 𝑢(𝑡)
𝑋 𝑗𝜔 𝑎 + 𝑗𝜔
Example 1

31
 32
Table 1: FOURIER Transform Properties
Table 2 BASIC FOURIER TRANSFORM PAIRS (Selected)

33