Signal Processing I

(23ECE112)
Ms. Latha S
Assistant Professor
Room no B-206, ECE department
Email id: s_latha@[Link]
 Complex exponential signals

• 𝑥 𝑡 = 𝐶𝑒 𝑎𝑡 where C and a are constants and are possibly complex valued.

• If C and a are real values, then it is the same as the real-valued exponential signal that you have
studied.
 Complex exponential signals

• 𝑥 𝑡 = 𝐶𝑒 𝑎𝑡 Consider the scenario where a is purely imaginary (𝑗𝜔0 ) and C=1

• Then 𝑥 𝑡 = 𝑒 𝑗𝜔0𝑡

• By Euler’s relation 𝑒 𝑗𝜃 = cos 𝜃 + 𝑗 sin(𝜃)

• Then 𝑥 𝑡 = 𝑒 𝑗𝜔0𝑡 = cos 𝜔0 𝑡 + 𝑗 sin 𝜔0 𝑡 = 𝑥 + 𝑗𝑦

• For any value of t, we will get a complex valued amplitude.

• Complex exponential signal is a periodic signal. What is its fundamental time period?

• If it is a periodic signal with period T, 𝑥 𝑡 + 𝑇 = 𝑥(𝑡)

 Complex exponential signals
2𝜋
• Fundamental period 𝑇0 = and 𝜔0 is called angular frequency and is measured in rad/s.
𝜔0

• What about 𝑥 𝑡 = 𝑒 −𝑗𝜔0𝑡 ? Is it periodic

• Then 𝑥 𝑡 = 𝑒 −𝑗𝜔0𝑡 = cos 𝜔0 𝑡 − 𝑗 sin 𝜔0 𝑡 = 𝑥 − 𝑗𝑦

2𝜋
• It is also periodic with period 𝑇0 = . Here 𝜔0 is negative
𝜔0
 Complex exponential signals
• 𝑥 𝑡 = 𝐶𝑒 𝑎𝑡

• C introduces an amplitude scaling and does not affect the shape of the signal. It could be real or
complex.

• A more generic form of complex exponential signal is 𝑥 𝑡 = 𝐶𝑒 𝑗(𝜔0𝑡+𝜑) where 𝜑 is denoted as the
phase.

• The signal can be rewritten 𝑥 𝑡 = 𝐶𝑒 𝑗𝜑 𝑒 𝑗𝜔0 𝑡

• The value of phase determines the amplitude of the signal at t=0.

• 𝑥 𝑡 = 𝐶𝑒 𝑗𝜑 𝑒 𝑗𝜔0 𝑡 = 𝐶 cos(𝜔0 𝑡 + 𝜑) + 𝐶 sin(𝜔0 𝑡 + 𝜑) with C as real

• The real part of 𝐶𝑒 𝑗(𝜔0𝑡+𝜑) with C as real is 𝑦1 𝑡 = 𝐶 cos(𝜔0 𝑡 + 𝜑) which is cosine signal.

• The imaginary part of 𝐶𝑒 𝑗(𝜔0𝑡+𝜑) with C as real is 𝑦2 𝑡 = 𝐶 sin(𝜔0 𝑡 + 𝜑) which is sine signal.
 Cosine and sine signals

• ℜ𝑒 𝐶𝑒 𝑗 𝜔0 𝑡+𝜑 = 𝐶 ℜ𝑒 cos 𝜔0 𝑡 + 𝜑 + 𝑗 sin 𝜔0 𝑡 + 𝜑 = 𝐶 cos 𝜔0 𝑡 + 𝜑

• Similarly, 𝐼𝑚 𝐶𝑒 𝑗 𝜔0 𝑡+𝜑 = 𝐶 sin(𝜔0 𝑡 + 𝜑)

2𝜋
• We know that cos 𝜔0 𝑡 + 𝜑 and 𝑠𝑖𝑛 𝜔0 𝑡 + 𝜑 are periodic signals with period 𝑇0 =
𝜔0
Cosine and sine signal

The value of φ does not affect the fundamental period of the signal as it
provides only a time delay.
 Cosine and sine signal

𝑒 𝑗𝜃 +𝑒 −𝑗𝜃 𝑒 𝑗𝜃 −𝑒 −𝑗𝜃
• We also know that cos 𝜃 = and sin 𝜃 =
2 2𝑗

𝑒 𝑗(𝜔0𝑡+𝜑) +𝑒 −𝑗(𝜔0𝑡+𝜑) 𝐶
• x t = C cos 𝜔0 𝑡 + 𝜑 = 𝐶( ) = 𝑒 𝑗𝜑 𝑒 𝑗𝜔0𝑡 + 𝑒 −𝑗𝜑 𝑒 −𝑗𝜔0 𝑡
2 2

𝑒 𝑗𝜔0𝑡+𝜑 −𝑒 −𝑗(𝜔0𝑡+𝜑) 𝐶
• Similarly for x t = C sin 𝜔0 𝑡 + 𝜑 = 𝐶( ) = 𝑒 𝑗𝜑 𝑒 𝑗𝜔0 𝑡 − 𝑒 −𝑗𝜑 𝑒 −𝑗𝜔0𝑡
2𝑗 2𝑗

• The frequency of the sinusoid signal (in Hz) 𝑓0 is related to 𝜔0 as 𝜔0 = 2π𝑓0 .

2𝜋 1
• But 𝜔0 = . Hence 𝑓0 =
𝑇0 𝑇0
Cosine and sine signal
 Practice problem
• Find the maximum amplitude, angular frequency, time period, frequency and phase of the
following signals :
𝜋
𝑥 𝑡 = 5 cos 𝑡 , 𝑦 𝑡 = 10 sin 200𝜋𝑡 , 𝑧 𝑡 = 2 cos(6000𝜋𝑡 + )
2
 Energy and power of complex exponential
signal
• Consider 𝑥 𝑡 = 𝑒 𝑗𝜔0 𝑡 , What would be the energy of the signal for a fundamental time period duration?

𝑇 𝑇 𝑇
• 𝐸𝑇0 = ‫׬‬0 0 |𝑥 𝑡 |2 𝑑𝑡 = ‫׬‬0 0 |𝑒 𝑗𝜔0 𝑡 |2 𝑑𝑡 = ‫׬‬0 0 1 𝑑𝑡 = 𝑇0

• What will be 𝐸∞ if energy of one time period is 𝑇0?

• It will be ∞.

• What is the average power over one fundamental time period duration?

1 𝑇0 1 𝑇0 𝑗𝜔0 𝑡 2 1 𝑇0
• 𝑃𝑇0 = ‫𝑥| ׬‬ 𝑡 |2 𝑑𝑡 = ‫𝑒| ׬‬ | 𝑑𝑡 = ‫ ׬‬1 𝑑𝑡 = 1
𝑇0 0 𝑇0 0 𝑇0 0

• What is the average power over two fundamental time period duration?

1 2𝑇0
• 𝑃2𝑇0 = 2𝑇 ‫׬‬0 |𝑥 𝑡 |2 𝑑𝑡 = 1
0

• Therefore, the total average power over infinite time duration 𝑃∞ = 1. Hence 𝑥 𝑡 = 𝑒 𝑗𝜔0 𝑡 is a power signal.
 Summary
• 𝑥 𝑡 = 𝑒 𝑗𝜔0 𝑡 = cos 𝜔0 𝑡 + 𝑗 sin 𝜔0 𝑡 = 𝑥 + 𝑗𝑦
• 𝑥 𝑡 = 𝑒 −𝑗𝜔0 𝑡 = cos 𝜔0 𝑡 − 𝑗 sin 𝜔0 𝑡 = 𝑥 − 𝑗𝑦
2𝜋
• Both are periodic with fundamental period 𝑇0 =
𝜔0
• Real part of x(t) =R{x(t)}= cos 𝜔0 𝑡
• Imaginary part of x(t) = Im{x(t)} = sin 𝜔0 𝑡

• 𝑥 𝑡 = 𝐶𝑒 𝑗(𝜔0 𝑡+𝜑) = 𝐶 cos(𝜔0 𝑡 + 𝜑) + 𝐶 sin(𝜔0 𝑡 + 𝜑) where is a C as real constant

2𝜋
• It is periodic with fundamental period 𝑇0 =
𝜔0
• Real part of x(t) =R{x(t)}=𝐶 cos(𝜔0 𝑡 + 𝜑)
• Imaginary part of x(t) = Im{x(t)} =𝐶 sin(𝜔0 𝑡 + 𝜑)

1 𝑇0 1 𝑇0 𝑗𝜔0 𝑡 2 1 𝑇0
• The average power is 𝑃= ‫𝑥| ׬‬ 𝑡 |2 𝑑𝑡 = ‫𝑒| ׬‬ | 𝑑𝑡 = ‫ ׬‬1 𝑑𝑡 = 1
𝑇0 0 𝑇0 0 𝑇0 0
 Harmonics of complex exponential signal
2𝜋
• Let 𝑥1 𝑡 = 𝑒 ±𝑗𝜔0 𝑡 which is periodic with fundamental period 𝑇0 = , fundamental angular
𝜔0
frequency 𝝎𝟎 rad/sec and fundamental frequency f0 Hz.

• Consider that 𝑥2 𝑡 = 𝑒 ±𝑗2𝜔0 𝑇 = cos 2𝜔0 𝑇 ± 𝑗 sin 2𝜔0 𝑇 which is periodic with fundamental period
2𝜋 𝑇0
𝑇0 = 2𝜔 = and fundamental angular frequency 𝟐𝝎𝟎 rad/secand fundamental frequency 2f0 Hz.
0 2

• Similarly 𝑥3 𝑡 = 𝑒 ±𝑗3𝜔0 𝑇 = cos 3𝜔0 𝑇 ± 𝑗 sin 3𝜔0 𝑇 which is periodic with fundamental period
2𝜋 𝑇0
𝑇0 = 3𝜔 = and fundamental angular frequency 𝟑𝝎𝟎 rad/sec and fundamental frequency 3f0 Hz.
0 3

• In general, any signal x(t) of the form 𝑥𝑘 𝑡 = 𝑒 ±𝑗𝑘𝜔0 𝑇 = cos 𝑘𝜔0 𝑇 ± 𝑗 sin 𝑘𝜔0 𝑇 which is periodic
2𝜋 𝑇0
with fundamental period 𝑇0 = 3𝜔 = 𝑘
and fundamental angular fundamental frequency 𝒌𝝎𝟎 rad/sec
0

and frequency k f0 Hz.

 Generalized complex exponential signal Cont…

• Thus, the set of complex exponential signals denoted by 𝝋𝒌 𝒕 = 𝒆𝒋𝒌𝝎𝟎 𝒕 for 𝒌 = 𝟎, ±𝟏, ±𝟐, … are
all periodic and their angular frequency and frequency are integer multiples of 𝜔0 and 𝑓0
respectively.

• This set of complex exponential signals are said to be harmonically related. The frequencies,
2𝜔0 , 3𝜔0 ,4𝜔0 are called the harmonics of the frequency 𝜔0 .

• This set of complex exponential signals can be used to represent other signals (Fourier
Representation).
 Generalized complex exponential signal

• 𝑥 𝑡 = 𝐶𝑒 𝑎𝑡 where C and a are complex valued.

• Generally, 𝑎 = 𝑟 + 𝑗𝜔0 .

• We can write the value C in polar form as 𝐶 = 𝐶 𝑒 𝑗𝜑 𝑤ℎ𝑒𝑟𝑒 𝜑 = ∠𝐶

• Then 𝑥 𝑡 = 𝐶𝑒 𝑎𝑡 = |𝐶|𝑒 𝑗𝜑 𝑒 (𝑟+𝑗𝜔0 )𝑡 = 𝐶 𝑒 𝑟𝑡 𝑒 𝑗(𝜔0𝑡+𝜑)

• Now how does this signal look like?

• Since the signal is complex, we will try to plot the real and imaginary parts separately.
 Generalized complex exponential signal
• ℜ𝑒 𝐶 𝑒 𝑟𝑡 𝑒 𝑗(𝜔0𝑡+𝜑) = 𝐶 𝑒 𝑟𝑡 cos(𝜔0 𝑡 + 𝜑)

• Similarly, 𝐼𝑚 𝐶 𝑒 𝑟𝑡 𝑒 𝑗(𝜔0𝑡+𝜑) = 𝐶 𝑒 𝑟𝑡 sin(𝜔0 𝑡 + 𝜑)

Real part is the product of real valued exponential signal with cosine
signal. For r<0 we get a decaying/damped cosine signal.
 Complex exponential signal

• 𝑥 𝑡 = 𝑒 𝑗𝜔0𝑡 , What happens when 𝜔0 =0?

• Then 𝑥 𝑡 = 1 ∀ t also known as constant signal/ DC signal

 Discrete time complex exponential signal
• 𝑥 𝑛 = 𝐶𝛼 𝑛 , where C and 𝛼 are constants and are complex
numbers in general.

• If we write 𝛼 = 𝑒 𝛽 Then we have 𝑥 𝑛 = 𝐶𝑒 𝛽𝑛 .

• If C and 𝛼 are real constants, then the equation for x[n] is similar
to the discrete time real-valued exponential signal.

• If |𝛼|>1 then x[n] is a growing exponential signal and if |𝛼|<1 then

x[n] is a decaying exponential signal

• If 𝛼 value is negative, then the values of x[n] alternates between

positive and negative values.
 Discrete time sinusoidal signal
• 𝑥 𝑛 = 𝐶𝑒 𝛽𝑛 , consider 𝛽 to be purely imaginary (say 𝑗𝛺0 ) and C=1, then we have 𝑥 𝑛 = 𝑒 𝑗𝛺0𝑛
and is the expression for discrete time complex exponential signal
 Discrete time complex sinusoidal signal
• Now let us consider C to be complex. We can write the value C in polar form as 𝐶 = 𝐶 𝑒 𝑗𝜑

• 𝑥 𝑛 = 𝐶𝑒 𝑗𝛺0𝑛 = |𝐶|𝑒 𝑗𝜑 𝑒 𝑗𝜔0𝑛 = 𝐶 𝑒 𝑗(𝛺0𝑛+𝜑)

• Let A=|C| we have ℜ𝑒 𝐴 𝑒 𝑗(𝛺0𝑛+𝜑) = 𝐴 cos(𝛺0 𝑛 + 𝜑) and 𝐼𝑚 𝐴 𝑒 𝑗(𝛺0𝑛+𝜑) = 𝐴 sin(𝛺0 𝑛 + 𝜑)

• For any value of n, |𝑒 𝑗𝛺0𝑛 |2 = 1. Therefore, the energy of one sample of discrete time complex
exponential signal is 1.

• Therefore, total energy 𝐸∞ = ∞.

1
• Average power over N samples is 1 + 1 + 1 + ⋯ + 1 = 1.
𝑁

• Therefore 𝑃∞ = 1 N times

• Hence it is a power signal.

 Generalized discrete time complex exponential
signal

• 𝑥 𝑛 = 𝐶𝛼 𝑛 , where C and 𝛼 are complex in

nature. Hence 𝐶 = 𝐶 𝑒 𝑗𝜑 , 𝛼 = 𝛼 𝑒 𝑗𝜔0 .

• 𝑥 𝑛 = 𝐶 𝑒 𝑗𝜑 𝛼 𝑛 𝑒 𝑗𝜔0𝑛 = 𝐶 𝛼 𝑛

𝑒 𝑗(𝜔0𝑛+𝜑) .

• If 𝛼 <1 then the signal x[n] is a damped

sinusoid signal, If 𝛼 >1 then the signal x[n] is a
growing sinusoid signal.

• If 𝛼 =1 then the signal x[n] is a pure sinusoid

signal with constant peak amplitude.
 Angular frequency of discrete time complex
exponential
• Let us consider 𝑥1 𝑛 = 𝑒 𝑗𝛺0𝑛 and 𝑥2 𝑛 = 𝑒 𝑗𝛺1𝑛 where 𝜔1 = 𝛺0 + 2𝜋

• Then 𝑥2 𝑛 = 𝑒 𝑗(𝛺0+2𝜋)𝑛 = 𝑒 𝑗𝛺0𝑛 𝑒 𝑗2𝜋𝑛 = 𝑒 𝑗𝛺0𝑛 = 𝑥1 𝑛

• Therefore 𝑥1 𝑛 and 𝑥2 𝑛 are the same and we will not be able to differentiate the discrete
complex exponentials of angular frequency 𝛺0 and 𝛺0 +2𝜋.

• This is true for any integer multiples of 2𝜋. 𝑒 𝑗(𝛺0±2𝜋 𝑘)𝑛 = 𝑒 𝑗𝛺0𝑛

• In other words, discrete frequencies can be distinctly specified only in range of length 2𝜋.
 Angular frequency of the discrete time complex
exponential signal

• Note that this range of 2𝜋 can be considered anywhere in the 𝜔0 axis.

• Whatever frequencies that appear in this interval will repeat in the adjacent intervals of length 2𝜋.

• We generally consider the range as −𝜋, 𝜋 or 0,2𝜋

• We can increase the frequency of oscillation from 𝜔0 = 0 only up to the frequency 𝜔0 = 2𝜋.

• At 𝜔0 = 2𝜋, we are going to see the same frequency as 𝜔0 = 0

 Periodicity of the complex exponential signal
• Considering the range of angular frequencies as −𝜋, 𝜋 , what is the largest frequency possible ?

• How does it look like?

• 𝑥 𝑛 = 𝑒 𝑗𝜋𝑛 = (−1)𝑛

• However, for discrete time complex exponential signal this may not
be true always.

• 𝑒 𝑗𝜔0 (𝑛+𝑁) = 𝑒 𝑗𝜔0 𝑛

• 𝑒 𝑗𝜔0 𝑛 𝑒 𝑗𝜔0 𝑁 = 𝑒 𝑗𝜔0 𝑛
• This happens when 𝑒 𝑗𝜔0 𝑁 = 1
 Periodicity of complex exponential signal

• Or in other words 𝜔0 𝑁 = 2𝜋𝑘 where k is an integer

2𝜋𝑘
• 𝑁=
𝛺0

• For discrete time signal N must be an integer

2𝜋
• Therefore, for 𝑒 𝑗𝜔0 𝑛 to be periodic signal, the ratio must be rational number.
𝛺0
Differences between discrete time and continuous
time complex exponential signals
𝑒 𝑗𝜔0𝑡 𝑒 𝑗𝛺0𝑛

Distinct signals for distinct values of 𝜔0 Identical signals if 𝜔0 is separated by integer

multiples of 2𝜋
Periodic for any choice of 𝜔0 𝛺0
Periodic only if is a rational number.
2𝜋

Angular frequency is 𝜔0 𝛺0
Angular frequency is .
𝑘

2𝜋 2𝜋𝑘
𝑇0 = 𝑁0 =
𝜔0 𝛺0
 Discrete time complex exponential signal

• If 𝛺0 = 0 then 𝑒 𝑗 𝛺0𝑛 = 1.

• In the continuous time case, the unit of angular frequency was rad/s. What is the unit in the discrete time
scenario?

2𝜋 (𝑟𝑎𝑑𝑖𝑎𝑛𝑠)
• Angular frequency =
𝑁 (𝑖𝑛𝑡𝑒𝑔𝑒𝑟)

• Unit of angular frequency for discrete time case is rad not rad/s.

• We know that 𝑒 𝑗𝜔0𝑛 = cos 𝛺0 𝑛 + 𝑗 sin 𝛺0 𝑛. So, all the discussions that we had regarding periodicity and
frequency are applicable for discrete time sinusoidal signals also.
 Practice problem

2𝜋
• Is 𝑥 𝑛 = cos 𝑛 a periodic signal? If yes, what is the fundamental period and angular frequency?
12

2𝜋 𝜋
• cos 𝜔0 𝑛 = cos 𝑛 𝛺0 =
12 6

2𝜋
• 𝛺0
= 12 is a rational number. So 𝑥 𝑛 is periodic

2𝜋𝑘 2𝜋𝑘
• 𝑁= 𝛺0
= 𝜋 = 12𝑘
6

• Assuming k=1 we get N=12.

𝛺0 𝜋
• Angular frequency = =
𝑘 6
 Practice problem
8𝜋
• Is 𝑥 𝑛 = s𝑖𝑛 𝑛 a periodic signal? If yes, what is the fundamental period and angular frequency?
31

8𝜋 8𝜋
• sin 𝛺0 𝑛 = sin 𝑛 𝛺0 =
31 31

2𝜋 31
• = is a rational number. So 𝑥 𝑛 is periodic
𝜔0 4

2𝜋𝑘 2𝜋𝑘 31
• 𝑁= = 8𝜋 = 𝑘
𝛺0 4
31

• Assuming k=4 we get N=31.

𝛺0 2𝜋
• Angular frequency = 𝑘
= 31
 Practice problem

𝑛
• Is 𝑥 𝑛 = cos a periodic signal? If yes, what is the fundamental period and angular frequency?
6

𝑛 1
• cos 𝛺0 𝑛 = cos 𝛺0 =
6 6

2𝜋
• = 12𝜋 is not a rational number. So 𝑥 𝑛 is not a periodic signal
𝛺0
 Practice Problems
• Check if the following signals are periodic. If yes, find the fundamental period and angular
frequency
3𝜋
• 𝑥 𝑛 = 5 𝑐𝑜𝑠 𝑛
4
𝜋
• 𝑦 𝑛 = 𝑠𝑖𝑛 2𝜋𝑛 +
4
𝜋
• 𝑟 𝑛 = 6 𝑠𝑖𝑛 𝑛
14