Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

Module 1: Band-pass Signal to Equivalent Low-pass
Hilbert Transform, Pre-envelopes, Complex envelopes, Canonical representation of band-pass
signals, Complex low pass representation of band-pass systems, Complex representation of band
pass signals and systems (Text 1: 2.8, 2.9, 2.10, 2.11, 2.12, 2.13).
Line codes: Unipolar, Polar, Bipolar (AMI) and Manchester code and their power spectral densities
(Text 1: Ch 6.10). Overview of HDB3, B3ZS, B6ZS (Ref. 1: 7.2) L1, L2,L3

Hilbert Transform:
1. Define Hilbert Transform. List the properties of the Hilbert Transform.
June/July.2018, 15EC61, 04 Marks
2. Define Hilbert transform. What are its applications. Prove that a signal g(t) and its Hilbert transform ĝ(t)
are orthogonal over the entire time interval (–∞, ∞). Dec.2018/Jan.2019, 15EC61, 05 Marks
 Fourier transform is useful for evaluating the frequency content of an energy signal or power signal. It
provides the mathematical basic for analyzing and designing frequency selective filters for separation of
signals based on frequency content.
 Another method of separating the signals is based on phase selective, which uses the phase between the
signals to achieve the desired separation. That is, a phase shift of ±900 .
 When the phase angles of all components of a given signal are shifted by ±900 , the resulting function
of time is known as the Hilbert Transform of the signal.
 Hilbert Transform is called a quadrature filter, because to emphasize its distinct property it provides a
phase shift of ±900 .

For a signal g(t) and its Fourier Transform G(f). The Hilbert Transform of g(t) denoted by ĝ(t), and is given
by
∞
1 g(τ)
ĝ(t) = ∫ dτ − − − − − −(1)
π t−τ
−∞
Hilbert transformation is a linear operation.
The inverse Hilbert transform to obtain the original signal g(t), linearly recovered from ĝ(t), and is given
by
∞
1 ĝ(τ)
g(t) = − ∫ dτ − − − − − (2)
π t−τ
−∞
The functions, g(t) and ĝ(t) are Hilbert Transform pairs. The below table gives the some Hilbert Transform
pairs:
Table: Hilbert Transform Pairs
Time Function: Hilbert Transform: Time Function: Hilbert Transform:
1 t − 1⁄2
m(t) cos(2πfc t) m(t) sin(2πfc t) rect(t) − ln | |
π t + 1⁄2
1
m(t) sin(2πfc t) −m(t) cos(2πfc t) δ(t)
πt
1 t
cos(2πfc t) sin(2πfc t)
1 + t2 1 + t2
1
sin(2πfc t) − cos(2πfc t) −πδ(t)
t
sin⁡(t) 1 − cos⁡(t)
t t

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 1

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

From Eqn (1), the definition of Hilbert Transform is interpreted as, the convolution of g(t) with the time
function 1⁄πt. From the convolution theorem, the convolution of two functions in time domain is equivalent
to multiplication of their Fourier transform in frequency domain.
For time function 1⁄πt, Fourier Transform pair is,
1
⇌ −jsgn(f)
πt
Where sgn(f) is the signum function, defined in frequency domain as,
1,⁡⁡⁡⁡⁡⁡⁡⁡⁡f > 0
sgn(f) = { 0,⁡⁡⁡⁡⁡⁡⁡⁡⁡f = 0
−1,⁡⁡⁡⁡⁡⁡⁡f < 0
Fourier transform of g(t) is G(f), then by applying duality property, the Fourier Transform Ĝ (f) of ĝ(t) is
given by,
̂ (f) = −j⁡sgn(f)⁡G(f) − − − − − (3)
G
From Eqn(3),
 For a given signal g(t) which is Fourier transformable, we may obtain the Fourier transform of its Hilbert
transform ĝ(t) by passing g(t) through a linear time invariant system whose frequency response is equal
to −j⁡sgn(f).

Fig: Hilbert Transformer

 This system is considered as it produces a phase shift of −900 for all positive frequencies in the input
signal and +900 for all negative frequencies as shown in below figure.

(a)
(b)
Fig: (a) Magnitude Response and (b) Phase Response of Hilbert transform

 The amplitudes of all frequency components are unaffected by the transmission through the device, such
an ideal system is referred as Hilbert Transformer or Quadrature Filter.

Applications of Hilbert Transform:

 It is used in the generation of single side band modulation to realize phase selectivity.
 It provides the mathematical basis for the representation of band pass signals.

Problem 1: Determine the Hilbert transform of the function x(t) = cos(2πfc t).
Solution: Given, x(t) = cos(2πfc t)
Fourier transform of x(t) is,
1
X(f) = [δ(f − fc ) + δ(f + fc )]
2

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 2

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

We know that,
1 j
̂
X(f) = −jsgn(f)X(f) = −jsgn(f) { [δ(f − fc ) + δ(f + fc )]} = − sgn(f)[δ(f − fc ) + δ(f + fc )]
2 2
1
̂
X(f) = [δ(f − fc ) + δ(f + fc )]
2j
This is the Fourier transform of sin(2πfc t), hence Hilbert transform of cos(2πfc t) is sin(2πfc t)

Problem 2: Obtain the Hilbert transform of the following: (i) x(t) = (cos 2πft + sin 2πft), (ii) x(t) =
e−j2πft
Solution:
(i) x(t) = cos(2πft) + sin(2πft)
π π
x̂(t) = cos (2πft + ) + sin (2πft + ) = sin(2πft) − cos(2πft)
2 2
−j2πft
(ii) x(t) = e = cos(2πft) − j sin(2πft)
π π
x̂(t) = cos (2πft + ) − j sin (2πft + ) = sin(2πft) − j cos(2πft) = −je−j2πft
2 2
Problem 3: Determine the Hilbert transform of sin(2π𝑓c t)
Solution: Let,
1
x(t) = sin(2πfc t) = [ej2πfc t − e−j2πfc t ]
2j
1
X(𝑓) = FT[x(t)] = [δ(𝑓 − 𝑓c ) − δ(𝑓 + 𝑓c )]
2j
We know that,
1
̂
X(𝑓) = −jsgn(𝑓)X(𝑓) = −jsgn(𝑓) [δ(𝑓 − 𝑓c ) − δ(𝑓 + 𝑓c )]
2j
1
= − [sgn(𝑓)δ(𝑓 − 𝑓c ) − sgn(𝑓)δ(𝑓 + 𝑓c )]
2
1,⁡⁡⁡⁡⁡⁡⁡⁡⁡𝑓 > 0
sgn(𝑓) = { 0,⁡⁡⁡⁡⁡⁡⁡⁡⁡𝑓 = 0
−1,⁡⁡⁡⁡⁡⁡⁡𝑓 < 0
δ(𝑓 − 𝑓c ) = 1; ⁡⁡when⁡𝑓 − 𝑓c = 0 ⇒ 𝑓 = 𝑓c ⁡⁡i. e. f > 0⁡⁡𝑎𝑛𝑑⁡sgn(𝑓) = 1
δ(𝑓 + 𝑓c ) = 1; ⁡⁡when⁡𝑓 + 𝑓c = 0 ⇒ 𝑓 = −𝑓c ⁡⁡i. e. f < 0⁡⁡𝑎𝑛𝑑⁡sgn(𝑓) = −1
1
̂
X(𝑓) = − [δ(𝑓 − 𝑓c ) + δ(𝑓 + 𝑓c )]
2
Apply inverse Fourier transform on both side;
1
x̂(𝑡) = − [ej2πfc t + e−j2πfc t ] = − cos(2πfc t)
2

Properties of Hilbert Transform:

The Hilbert transform differs from Fourier transform, it exclusively operates in time domain. It has its own
properties, some of which are listed below.

Property 1: A signal g(t) and its Hilbert transform ĝ(t) have the same magnitude response.
̂ (f)|
ie,. |G(f)| = |G
Since the Fourier transform of ĝ(t) is equal to −j⁡sgn(f) times the Fourier transform of g(t), and magnitude
of −j⁡sgn(f) is equal to 1 for all f, we can say g(t) and ĝ(t) have the same amplitude response.

Property 2: If ĝ(t) is the Hilbert transform of g(t), then the Hilbert transform of ĝ(t) is −g(t).
ie,. agr[G(f)] = −agr[Ĝ (f)]

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 3

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

Hilbert transform is obtained by passing g(t) through a linear two port device with a transfer function
−j⁡sgn(f). A twice Hilbert transform is equivalent to passing g(t) through cascade of a linear two port
devices. The overall transfer function of a cascaded devices is given by,
[−j⁡sgn(f)]2 = −1,⁡⁡⁡⁡for⁡all⁡f
Hence we can say, agr[G(f)] = −agr[G ̂ (f)]

Property 3: A signal g(t) and its Hilbert transform ĝ(t) are orthogonal over the entire time interval (–∞,
∞).
In mathematical terms, the orthogonality of g(t) and ĝ(t) is described by
∞

∫ g(t) ĝ(t)dt = 0
−∞
Proof:
We know that,
g(t) ⇌ ĝ(t)⁡⁡⁡⁡then⁡⁡⁡⁡ĝ(t) ⇌ −g(t)
In frequency domain,
̂ (f)⁡⁡⁡⁡then⁡⁡⁡⁡G
G(f) ⇌ G ̂ (f) ⇌ G(−f)
∞ ∞ ∞ ∞
̂ (f) G(−f)⁡df = ∫ −j⁡sgn(f)⁡G(f) G(−f)⁡df = −j ∫ ⁡sgn(f)⁡G(f) G∗ (f)⁡df
∫ g(t) ĝ(t)dt = ∫ G
−∞ −∞ −∞ −∞
Where, G∗ (f) = G(−f) for real valued signal.
∞ ∞

∫ g(t) ĝ(t)dt = −j ∫ ⁡sgn(f)⁡|G(f)|2 df

−∞ −∞
The right hand side of the above equation is and integration of odd function, because it is a product of the
odd function sgn(f) and even function |G(f)|2. Hence its integration is zero.
∞

Therefore, ∫ g(t) ĝ(t)dt = 0

−∞
Hence proved.

Example: Hilbert transform of Low-Pass signal.

Consider the below figure, Fig(a) is the Fourier transform of low-pass signal g(t), whose frequency content
̂ (f) is
extends from – W to W. Hilbert transform to this low-pass signal is ĝ(t), and its Fourier transform G
shown in Fig(b).

(a)
(b)
Fig: (a) Spectrum of signal g(t) and (b) Spectrum of Hilbert transform 𝐠̂(𝐭)

Form the figure, the frequency content of Fourier transformable signal can be radically changed as a result
of Hilbert transformation.

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 4

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

Pre-envelopes:
 Hilbert transform is defined for both positive and negative frequencies. The basic question is, how can
we modify the frequency content of a real-valued signal g(t) such that all negative frequency content are
completely eliminated.
 Solution to this fundamental question lies in the idea of complex-valued signal called pre-envelope of
g(t) and is defined as,

g + (t) = g(t) + j⁡ĝ(t) − − − −(1)

Where ĝ(t) is the Hilbert transform of g(t).
̂ + (f)
From the above equation, g(t) is real part and ĝ(t) is imaginary part of the pre-envelope g + (t). Let G
denote the Fourier transform of g + (t) and is given by,
̂ + (f) = G(f) + j[−j⁡sgn(f)⁡G(f)] = G(f) + sgn(f)G(f)
G
Using the definition of sgn(f), we can write the above equation as,
2G(f),⁡⁡⁡⁡⁡⁡⁡⁡⁡f > 0
̂ + (f) = { G(0),⁡⁡⁡⁡⁡⁡⁡⁡⁡f = 0 − − − − − (2)
G
0⁡,⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡f < 0
Where, G(0) is the value of G(f) at the origin f = 0.
The above equation clearly shows that the pre-envelope of the signal g(t) has no frequency contents for all
negative frequencies, as illustrated in below figure.

(a)
(b)
Fig: (a) Spectrum of low-pass signal g(t) and (b) Spectrum of pre-envelope 𝐠 + (𝐭)
For a given signal g(t), pre-envelope g + (t) can be obtained in two methods:
1. Time-domain procedure: For a given g(t), first compute Hilbert transform ĝ(t), then pre-envelope
g + (t) is obtained by using Eqn(1).
2. Frequency-domain procedure: For a given g(t), first compute Fourier transform G(f), then Fourier
transform of pre-envelope G ̂ + (f) is obtained by using Eqn(2), and finally evaluate the Inverse Fourier
transform of G ̂ + (f) to obtain pre-envelope g + (t).
∞

g + (t) = 2 ∫ G(f)ej2πft df − − − −(3)

0
Eqn (1) defines the pre-envelope g+ (t) for positive frequencies, similarly we can define pre-envelope for
negative frequencies as,
g − (t) = g(t) − j⁡ĝ(t) − − − − − (4)
̂ − (f) denote the Fourier transform of g − (t) and is given by,
Let G
̂ − (f) = G(f) − j[−j⁡sgn(f)⁡G(f)] = G(f) − sgn(f)G(f)
G
Using the definition of sgn(f), we can write the above equation as,
0,⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡f > 0
̂
G− = { G(0),⁡⁡⁡⁡⁡⁡⁡⁡⁡f = 0 − − − −(5)
(f)
2G(f)⁡,⁡⁡⁡⁡⁡⁡⁡⁡f < 0
This shows that the pre-envelope g − (t) of a signal has no frequency content for all positive frequencies.

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 5

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

From equations (1) & (4), the two pre-envelopes g + (t) and g − (t) are simply the complex conjugate of each
other.
g − (t) = g ∗+ (t)
The spectrum of both the pre-envelopes g + (t): for all positive frequencies and g − (t): for all negative
frequencies are non-zero only. Thus the pre-envelopes g + (t) and g − (t) constitute a complementary pair of
complex-valued signals. Also sum of g + (t) and g − (t) is exactly equal to twice the original signal g(t).

From Eqn(2), for a given real-valued signal, the pre-envelope g + (t) is uniquely defined by the spectral
content for positive frequencies. Similarly, from Eqn(5), the pre-envelope g − (t) is uniquely defined by the
spectral content for negative frequencies.

Example: Pre-envelopes of Low-Pass Signal

The below figure illustrate the application of Hilbert transformation to a low-pass signal.

(a)
(b) (c)
Fig: (a) Spectrum of Low-Pass Signal, (b) Spectrum of the pre-envelope 𝐠 + (𝐭) and
(a) Spectrum of the pre-envelope 𝐠 − (𝐭)

From figure, the spectrum of g(t) is defined for −W ≤ f ≤ W, the spectral content of g + (t) is confined
entirely to 0≤ f ≤ W, and the spectral content of g − (t) is confined entirely to −W ≤ f ≤ 0.

Complex Envelopes of Band-Pass Signals:

1. Define pre-envelope of a real valued signal. Given a band pass signal s(t), sketch the amplitude spectral
of signal s(t), pre-envelope s+ (t) and complex envelope s̃(t). June/July.2018, 15EC61, 04 Marks

Let the band-pass signal denoted by s(t) and its Fourier transform is denoted by S(f). The Fourier transform
S(f) is essentially confined to a band of frequencies of total extent 2W, centered about some frequency ±fc .
Where fc is the carrier frequency. In majority of communication signals, the bandwidth 2W is small
compared to carrier frequency fc , hence we can refer the signal s(t) as a narrow band signal.

The pre-envelope of the narrow band signal s(t) is given by,

s+ (t) = s̃(t)ej2fc t − − − − − (1)
Where, s̃(t) is the complex envelope of the band-pass signal s(t). The below figure illustrate the magnitude
spectrum of band-pass signal, pre-envelope s+ (t) and complex envelope s̃(t).

(a)

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 6

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

(b) (c)
Fig: (a) Magnitude Spectrum of Band-Pass Signal s(t), (b) Magnitude Spectrum of pre-
envelope 𝐬+ (𝐭) and (c) Magnitude Spectrum of Complex Envelope 𝐬̃(𝐭)

From Fig(b), the spectrum of s+ (t) is limited to the positive frequency band fc − W ≤ f ≤ fc + W. Then
by applying the frequency-shifting property of the Fourier transform to Eqn(1), the spectrum of the complex
envelope s̃(t) is correspondingly limited to −W ≤ f ≤ +W and centered at the origin, illustrated in Fig (c).

The complex envelope s̃(t) of the band-pass signal s(t) is a complex low-pass signal. The essence of the
mapping from the band-pass signal s(t) to the complex low-pass signal s̃(t) are:
1. Information content of modulated signal s(t) is fully preserved in the complex envelope s̃(t).
2. Analysis of band-pass signal s(t) is complicated by the presence of the carrier frequency fc , the complex
envelope s̃(t) dispenses with fc making the analysis simpler.
3. The use of s̃(t) requires having to handle complex notations.

1. Determine the pre-envelope and complex envelope of the RF pulse defined by x(t) =
t
A⁡rect (T) cos(2πfc t). Dec.2018/Jan.2019, 15EC61, 06 Marks
Solution:
First find the Fourier transform of a given signal,
We know that,
1 j2πf t
cos(2πfc t) = [e c + e−j2πfc t ]
2
Therefore, the given is
t 1 A t A t
x(t) = A⁡rect ( ) [ej2πfc t + e−j2πfc t ] = ⁡rect ( ) ej2πfc t + ⁡rect ( ) e−j2πfc t
T 2 2 T 2 T
Applying Fourier transform along with the frequency shift property,
AT AT AT
X(f) = sinc[T(f − fc )] + sinc[T(f + fc )] = {sinc[T(f − fc )] + sinc[T(f + fc )]}
2 2 2
Note:
1. Frequency Shift Property: g(t)e−j2πf0 t ⇌ G(f − f0 )
t
2. Fourier Transform Pair: rect (T) ⇌ T⁡sinc(Tf)
Assume that, fc T ≫ 1 then x(t) is considered as narrow band signal. Then the result can be approximated
as,
AT
sinc[T(f − fc )]⁡⁡,⁡⁡⁡⁡f > 0
2
X(f) = 0⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡,⁡⁡⁡f = 0
AT
{ 2 sinc[T(f + fc )]⁡⁡,⁡⁡⁡⁡f < 0
Hence,
AT⁡sinc[T(f − fc )]⁡⁡,⁡⁡⁡⁡f > 0
X+ (f) = {
0⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡,⁡⁡⁡f ≤ 0

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 7

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

By taking inverse Fourier transform of X+ (f), we get pre-envelope as,
t
x+ (t) = A⁡rect ( ) ej2πfc t = x̃(t)⁡ej2πfc t
T
Therefore, the complex envelope is given by
t
x̃(t) = A⁡rect ( )
T
The natural envelope: a(t) = |x̃(t)|
t
∴ ⁡⁡a(t) = A⁡rect ( )
T
Canonical Representation of Band-Pass Signal:
1. Obtain the canonical representation of band pass signals. June/July.2018, 15EC61, 06 Marks
2. Express bandpass signal s(t) in canonical form. Also explain the scheme for deriving the in-phase
and quadrature components of the bandpass signal s(t). Dec.2018/Jan.2019, 15EC61, 06 Marks
From the definition of pre-envelope [s+ (t) = s(t) + j⁡ŝ(t)], the real part of the pre-envelope s+ (t) is equal
to the original band-pass signal s(t). Therefore, the band-pass s(t) can be expressed in-terms of its
corresponding complex envelope s̃(t) as,
s(t) = Re[s+ (t)] = Re[s̃(t)ej2fc t ] − − − − − (1)
Since s̃(t) is complex-valued quantity, it can be expressed in Cartesian form as,
s̃(t) = sI (t) + jsQ (t) − − − −(2)
(t) (t)
Where, sI and sQ are both real-valued low-pass functions.
Substitute Eqn(2) into Eqn(1),
s(t) = Re [[sI (t) + jsQ (t)][cos(2πfc t) + j sin(2πfc t)]]
s(t) = Re[sI (t) cos(2πfc t) + jsI (t) sin(2πfc t) + jsQ (t) cos(2πfc t) + j(jsQ (t)) sin(2πfc t)]
s(t) = Re[{sI (t) cos(2πfc t) − sQ (t) sin(2πfc t)} + j{sI (t) sin(2πfc t) − sQ (t) cos(2πfc t)}]
Therefore, the original band-pass signal s(t) can be expresses in canonical form or standard form as,
s(t) = sI (t) cos(2πfc t) − sQ (t) sin(2πfc t) − − − − − (3)
In the above equation, sI (t) as the in-phase component of the band-pass signal s(t) and sQ (t) as the
quadrature-phase component or quadrature component of the signal s(t).

From Eqn(1), the complex envelope s̃(t) may be pictured as a time-varying phasor positioned at the origin
of the (sI , sQ ) plane as shown in below fig(a). With time t varying continuously, the end of the phasor
moves about in the plane. Fig(b) depicts the phasor representation of the complex exponential ej2fc t .
Again from Eqn(1), the complex envelope s̃(t) is multiplied by the complex exponential ej2fc t . Hence, the
angles of these two phasors are added and their lengths are multiplied as shown in Fig(c). Also as shown
in Fig(c), the (sI , sQ ) phase rotating with an angular velocity 2πfc radians/sec. Hence, the phasor
representing the complex envelope s̃(t) moves in the (sI , sQ ) plane at same time the plane itself rotates
about the origin. Therefore, the original band-pass signal s(t) is the projection of this time-varying phasor
on a fixed line representing the real axis.

(a) (b)

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 8

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

(c)
Fig: Illustrating an interpretation of the complex envelope 𝐬̃(𝐭) and its multiplication by 𝐞𝐣𝟐𝐟𝐜𝐭

From Eqn(3), if cos(2πfc t), the multiplying factor of sI (t), is viewed as the reference sinusoidal carrier,
then the sin(2πfc t), the multiplying factor of sQ (t), is in phase quadrature with respect to cos(2πfc t).
Both the signals sI (t) and sQ (t) are low-pass signals limited to the band −W ≤ f ≤ W, they may
extracted from the band-pass signal s(t) using the following scheme shown in figure. It extracts the in-phase
and quadrature components, sI (t) and sQ (t), from the band-pass signal s(t). Hence, it can be referred as an
analyzer.

Fig: Deriving the in-phase and quadrature components of a band-pass signal

The below figure shows the reconstruction of band-pass signal s(t) from in-phase and quadrature
components, sI (t) and sQ (t). It can be referred as a synthesizer.

Fig: Reconstructing the band-pass signal from in-phase and quadrature components

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 9

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

 The above schemes are basic to the study of linear modulation schemes. Multiplication of the low-pass
in-phase component sI (t) by cos(2πfc t) and multiplication of quadrature component sQ (t) by
sin(2πfc t) represents linear form of modulation.
 The carrier frequency fc is larger than the low-pass bandwidth W, the resulting pass-band function s(t),
defined in Eqn(3) is referred as pass-band signal waveform. The mapping from sI (t) and sQ (t) combined
into s(t) is known as pass-band modulation.

Polar Representation of Band-Pass Signals:

The complex envelope s̃(t) of the band-pass signal s(t) is defined in polar form as,
s̃(t) = a(t)ejϕ(t)
Where a(t) & ϕ(t) are both real-valued low-pass functions.
Based on the polar representation, the original band-pass signal s(t) is defined as,
s(t) = a(t) cos[2πfc t + ϕ(t)]
We refer a(t) as the natural envelope or simply the envelope of the band-pass signal s(t) and ϕ(t) as the
phase of the signal.

Relationship Between Cartesian and Polar Representations of Band-Pass Signal:

The envelope a(t) and phase ϕ(t) are respectively related to the in-phase and quadrature components sI (t)
and sQ (t) as follows (from time-varying phasor representation)
a(t) = √sI2 (t) + sQ2 (t)
and
sQ (t)
ϕ(t) = tan−1 ( )
sI (t)
Conversely, we may write
sI (t) = a(t) cos[ϕ(t)]
And
sQ (t) = a(t) sin[ϕ(t)]
Hence, both in-phase and quadrature components of a band-pass signal contain amplitude and phase
information, both of which are uniquely defined for a prescribed phase ϕ(t), modulo 2π.

Complex Low-Pass Representation of Band-Pass Systems:

1. Derive the expression for the complex low pass representation of band pass systems.
June/July.2018, 15EC61, 08 Marks
2. Explain with relevant expressions, the procedure for computational analysis of a bandpass system
driven by bandpass signal. Dec.2018/Jan.2019, 15EC61, 06 Marks
We know that, how to handle the complex representation of band-pass signals, correspondingly we develop
a procedure for handling the representation of linear time-invariant band-pass systems. Specifically, the
analysis of band-pass systems is simplified by establishing an analogy or isomorphism between band-pass
and low-pass systems.

Consider a narrow band signal s(t) with its Fourier transform denoted by S(f). Assume that the spectrum of
the signal s(t) is limited to the frequencies within ±W Hz of the carrier frequency fc . Also assume that W <
fc .
Let the signal s(t) be applied to a linear time-invariant band-pass system with impulse response h(t) and
frequency response H(f). Again assume that the frequency response of the system is limited to frequencies
within ±B of the carrier frequency fc . The system bandwidth 2B is narrower than or equal to the input signal
bandwidth 2W.
Now represent the band-pass impulse response h(t) in terms of two quadrature components hI (t) and hQ (t),
by using the analogy to the representation of band-pass signals, we can express h(t) as,
h(t) = hI (t) cos(2πfc t) − hQ (t) sin(2πfc t) − − − −(1)

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 10

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

Correspondingly, the complex impulse response of the band-pass system is given by,
h̃(t) = hI (t) + jhQ (t) − − − − − (2)
Then, we may express h(t) in terms of h̃(t) as,
h(t) = Re[h̃(t)ej2πfc t ] − − − − − (3)
Note: hI (t), hQ (t) and h̃(t) are all low-pass functions, limited to the frequency band −B ≤ f ≤ B.
Now, we can determine the complex impulse response h̃(t) in terms of the in-phase and quadrature
components hI (t) and hQ (t) or from the band-pass frequency response H(f) as follows,
Let h̃∗ (t) be the complex conjugate of h̃(t), then Eqn(3) can be written in terms of h̃∗ (t) as
h(t) = Re[h̃∗ (t)e−j2πfct ] − − − −(4)
Adding Eqn(3) & Eqn(4),
2h(t) = h̃(t)ej2πfct + h̃∗ (t)e−j2πfc t − − − −(5)
Applying the Fourier transform on both sides of the Eqn(5) and using the complex-conjugation property of
the Fourier transform, we get
2H(f) = H ̃ (f − fc ) + H
̃ ∗ [−(f + fc )]
2H(f) = H ̃ (f − fc ) + H
̃ (−f − fc ) − − − −(6)
∗

Where, H(f) ⇌ h(t) and H ̃

̃ (f) ⇌ h(t)
Note: Complex conjugation property: FT{x(t)} = X(f) ⇒ FT{x ∗ (t)} = x ∗ (−f)

Eqn(6) satisfies the requirement H ∗ (f) = H(−f) for a real-valued impulse response h(t). Since H ̃ (f)
represents a low-pass frequency response limited to |f| ≤ B with B ≤ fc , we can write,
̃ (f − fc ) = 2H(f),⁡⁡⁡⁡f > 0 − − − − − (7)
H
Eqn(7) states that:
For a specified band-pass frequency response H(f), the corresponding complex low-pass frequency
response H̃ (f) is obtained by taking the part of H(f) defined for positive frequencies, shifting it to the origin
and scaling the amplitude by 2.

The complex frequency response H ̃ (f) can be decomposed into its in-phase and quadrature components as
follows,
̃ (f) = H
H ̃ I (f) + jH
̃ Q (f)
Where the in-phase component is defined as,
1
̃ I (f) = [H
H ̃ (f) + H ̃ ∗ (−f)]
2
And the quadrature component is defined as,
1
̃ Q (f) = [H
H ̃ (f) − H
̃ ∗ (−f)]
2j
Finally, the complex impulse response h̃(t) is obtained by taking the inverse Fourier transform of H
̃ (f).
∞

∴ ⁡ h̃(t) = ∫ H
̃ (f)ej2πft df
−∞
Complex Representation of Band-Pass Signals and Systems:
The band-pass signal s(t) in terms of complex envelope s̃(t) is,
s(t) = Re[s̃(t)ej2fc t ]
The impulse response h(t) of the band-pass system in terms of its complex impulse response h̃(t) is,
h(t) = Re[h̃(t)ej2πfc t ]
The above equations has common multiplying factor: ej2πfc t .
 In practical, this factor accounts for a sinusoidal carrier frequency (fc ), which facilitates transmission of
the modulated (band-pass) signal s(t) across a band-pass channel of mid-band frequency fc .
 In analytic terms, this exponential factor complicates the analysis of the band-pass system driven by the
modulated signal s(t).

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 11

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

 This analysis can be simplified through the combined use of complex low-pass equivalent
representations of both modulated signal s(t) and band-pass system characterized by the impulse
response h(t).
 The simplification can be carried out in the time domain or frequency domain.

Time Domain Procedure:

 With the complex representations of band-pass signal and systems, we can derive an analytically
efficient method for determining the output of a band-pass system driven by a corresponding band-pass
signal.
 Assume S(f) denotes the spectrum of input signal s(t) and H(f) denotes the frequency response of the
system, and both are centered around the same frequency fc .
 In practice, there is no need to consider a situation in which the carrier frequency of the input signal is
not aligned with the mid-band frequency of the band-pass system, since we have a freedom in choosing
the carrier or mid-band frequency.
 Thus, changing the carrier frequency of the input signal by an amount ∆fc , that corresponds to the factor
ej2π∆fc t in the complex envelope of the input signal or the complex impulse response of the band-pass
system. Still we can justify the assumption that S(f) and H(f) are both centered around the same carrier
frequency fc .

Let x(t) denotes the output signal of the band-pass system in response to the incoming band-pass signal s(t).
Clearly, x(t) is also a band-pass signal, so we can represent it in terms of its own low-pass complex envelope
x̃(t) as,
x(t) = Re[x̃(t)ej2πfc t ] − − − −(1)
The output signal x(t) is related to the input signal s(t) and impulse response h(t) by convolution integral
as,
∞

x(t) = ∫ h(τ)s(t − τ)dτ − − − −(2)

−∞
In terms of pre-envelopes, h(t) = Re[h+ (t)] and s(t) = Re[s+ (t)]. We can rewrite the Eqn(2) in terms of
the pre-envelopes s+ (t) and h+ (t) as,
∞

x(t) = ∫ Re[h+ (t)]Re[s+ (t − τ)]dτ − − − −(3)

−∞
By using the basic property of pre-envelopes that is described by the following relations:
∞ ∞
1
∫ Re[h+ (τ)]Re[s+ (τ)]dτ = Re [ ∫ h+ (τ)s+∗ (τ)dτ] − − − −(4)
2
−∞ −∞
From Fourier transform theory, using s(−τ) in place of s(τ) has the effect of removing the complex
conjugation on the RHS of Eqn(4).
By using the algebraic difference between the arguments s+ (t) in Eqn(4) and that of s+ (t − τ) in Eqn(3),
and using the relationship between the pre-envelopes and complex envelope of a band-pass signal, we write
the Eqn(3) in the equivalent form as,
∞
1
x(t) = Re [ ∫ h+ (τ)s+ (t − τ)dτ]
2
−∞
∞ ∞
1 1
x(t) = Re { ∫ h̃(τ)ej2πfc τ ⁡s̃(t − τ)ej2πfc (t−τ) dτ} = Re { ∫ ej2πfc τ ej2πfc t e−j2πfc τ h̃(τ)⁡s̃(t − τ)dτ}
2 2
−∞ −∞

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 12

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

∞
1
x(t) = Re { ∫ ej2πfc t h̃(τ)⁡s̃(t − τ)dτ} − − − −(5)
2
−∞
By comparing RHS of Eqn’s (1) & (5), for a large carrier frequency fc , the complex envelope x̃(t) of the
output signal is defined in terms of the complex envelope s̃(t) of the input signal and the complex impulse
response h̃(t) of the band-pass system as follows:
∞
1
x̃(t) = ∫ ej2πfc t h̃(t)⁡s̃(t − τ)dτ − − − (6)
2
−∞
This important relationship is the result of isomorphism between a band-pass function and the
corresponding complex low-pass function.

From the above equation, except the scaling factor 1⁄2, the complex envelope x̃(t) of the output of a band-
pass system is obtained by convolving the complex impulse response h̃(t) of the system with the complex
envelope s̃(t) of the input band-pass signal.

Fig: (a) Input-Output description of a band-pass system, (b) Complex low-pass equivalent
model of the band-pass system
Eqn: (6) can also be written as,
2x̃(t) = h̃(t) ∗ s̃(t)
Where, the symbol * indicates convolution.

The complex envelope s̃(t) of the input signal and the complex impulse response h̃(t) of the band-pass
systems are defined in terms of their respective in-phase and quadrature components as follows:
s̃(t) = sI (t) + jsQ (t)
̃h(t) = hI (t) + jhQ (t)
Substitute these equations into above equation, we get
2x̃(t) = [hI (t) + jhQ (t)] ∗ [sI (t) + jsQ (t)]
Since the convolution is distributive, above equation can be written as,
2x̃(t) = hI (t) ∗ sI (t) + jhI (t) ∗ sQ (t) + jhQ (t) ∗ sI (t) − hQ (t) ∗ sQ (t)
2x̃(t) = [hI (t) ∗ sI (t) − hQ (t) ∗ sQ (t)] + j[hI (t) ∗ sQ (t) + hQ (t) ∗ sI (t)] − − − (7)
Similarly, the complex envelope x̃(t) of the response can be defined in terms of in-phase and quadrature
components as,
x̃(t) = xI (t) + jxQ (t)⁡; ⁡⁡Multiply⁡both⁡side⁡by⁡2
2x̃(t) = 2xI (t) + j2xQ (t) − − − −(8)
By comparing the real and imaginary parts of equation’s (7) & (8), we get
In-phase component is defined by the relation,
2xI (t) = hI (t) ∗ sI (t) − hQ (t) ∗ sQ (t)
And quadrature component is defined by the relation,
2xQ (t) = hI (t) ∗ sQ (t) + hQ (t) ∗ sI (t)
The in-phase and quadrature components of the complex envelope x̃(t) of the system output are obtained
using the low-pass equivalent model as shown in below figure.

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 13

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

Fig: Block diagram illustrating the relationship between the in-phase and quadrature
components of the response of band-pass system and band-pass input signal

All the signals and impulse responses shown are real-valued low-pass functions, hence it is the time-domain
procedure for simplifying the analysis of band-pass systems driven by band-pass signals.

Frequency-Domain Procedure:
Time-domain expression is given by,
∞
1
x̃(t) = ∫ h̃(t)⁡s̃(t − τ)dτ
2
−∞
Also, we know that, from convolution thereom, convolution in time-domain is equivalent to multiplication
in frequency-domain, we get
1
̃
X(f) = H ̃ (f)S̃(f)
2
Where, s̃(t) ⇌ S̃(f), h̃(t) ⇌ H ̃ (f) and x̃(t) ⇌ ̃
X(f)
Assuming that H(f) is known, the frequency-domain procedure is summarized as below:
1. To determine H ̃ (f), use the equation H̃ (f − fc ) = 2H(f),⁡⁡⁡⁡f > 0
2. Express the input band-pass signal s(t) in canonical form: s(t) = sI (t) cos(2πfc t) −
sQ (t) sin(2πfc t), evaluate the complex envelope s̃(t) = sI (t) + jsQ (t). Then compute the Fourier
transform S̃(f) = FT{s̃(t)}.
3. Compute X ̃(f) = 1 H ̃ (f)S̃(f), which is the Fourier transform of the complex envelope x̃(t) of the
2
output signal x(t).
4. Compute the inverse Fourier transform of X ̃(f), yields x̃(t) = F −1 {X
̃(f)}
5. Then the desired output signal x(t) is obtained by, x(t) = Re[x̃(t)ej2πfc t ]

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 14

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

Line Codes:
The three basic waveform coding schemes are: PCM, DPCM, and DM. They differ from each other in
several ways: (i) transmission – bandwidth requirement, (ii) transmitter – receiver structural composition
and complexity, and (iii) quantization noise.
But all three of them have common need: line codes for electrical representation of the encoded binary
streams produced by their individual transmitters, to facilitate transmission of the binary streams across the
communication channel.
These waveforms are commonly known as digital data formats or their representation is called as line
coding. There are five different line codes, they are:
(i) Unipolar NRZ Signaling
(ii) Unipolar RZ Signaling
(iii) Polar NRZ Signaling
(iv) Bipolar RZ Signaling
(v) Split-Phase or Manchester Code

Unipolar:
In unipolar format, the waveform have a single polarity. The waveform can have +5V or +12V when high.
The waveform is simple on-off signaling.

Unipolar NRZ Signaling:

In this line code, symbol ‘1’ is represented by transmitting a pulse of amplitude ‘A’ for the complete
duration of the symbol, and symbol 0 is represented by switching off the pulse. i.e. for unipolar NRZ form,
If symbol ‘1’ is transmitted,
x(t) = A⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁡⁡(complete⁡interval)
If symbol ‘0’ is transmitted,
x(t) = 0⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁡⁡(complete⁡interval)
The pulse does not return to zero on its own. If symbol ‘0’ is to be transmitted, then pulse becomes zero.
Since there is no separation between the pulses, the receiver needs synchronization to detect unipolar NRZ
pulses.
Disadvantage: Waste of power due to the transmitted DC level which does not carry any information.

Unipolar RZ Signaling:
In this line code, symbol ‘1’ is represented by a rectangular pulse of amplitude ‘A’ for half-symbol width,
and symbol 0 is represented by transmitting no pulse. i.e. for unipolar RZ form,
If symbol ‘1’ is transmitted,
A⁡⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁄2 ⁡⁡⁡⁡(half⁡interval)
x(t) = {
0⁡⁡⁡⁡for⁡⁡⁡ Tb ⁄2 ≤ t ≤ Tb ⁡⁡(half⁡interval)
If symbol ‘0’ is transmitted,
x(t) = 0⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁡⁡(complete⁡interval)
Advantage: Presence of the delta function at f = 0, and it can be used for bit-timing recovery at the receiver.
Disadvantage: It requires 3 dB more power than polar RZ signaling for the same probability of symbol
error.

Polar (Bipolar) NRZ Signaling:

In this line code, symbols 1 and 0 are represented by transmitting a pulse of amplitudes +A and −A
respectively. That means, symbol ‘1’ is represented by positive voltage polarity and symbol ‘0’ is
represented by negative voltage polarity. i.e. for polar NRZ form,
If symbol ‘1’ is transmitted,
x(t) = +A⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁡⁡(complete⁡interval)
If symbol ‘0’ is transmitted,
x(t) = −A⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁡⁡(complete⁡interval)
Advantage: Easy to generate.

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 15

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

Disadvantage: The power spectrum of the signal is large near zero frequency.

Bipolar RZ Signaling:
It is also called as alternate mark inversion (AMI) signaling. In this line code, three amplitude levels are
used. Specifically positive and negative pulses of equal amplitudes are used alternatively for symbol 1, with
each pulse having a half symbol width. No pulse is used for symbol 0.
If symbol ‘1’ is transmitted,
+A⁡⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁄2 ⁡⁡⁡⁡(half⁡interval)⁡⁡⁡
x(t) = {
0⁡⁡⁡⁡⁡for⁡⁡⁡ Tb ⁄2 ≤ t ≤ Tb ⁡⁡(half⁡interval)
For next successive 1:
−A⁡⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁄2 ⁡⁡⁡⁡(half⁡interval)⁡⁡⁡⁡
x(t) = {
0⁡⁡⁡⁡⁡for⁡⁡⁡ Tb ⁄2 ≤ t ≤ Tb ⁡⁡(half⁡interval)
If symbol ‘0’ is transmitted,
x(t) = 0⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁡⁡(complete⁡interval)
Advantages: The power spectrum of the transmitted signal has no DC component and relatively
insignificant low-frequency components when symbols 1 & 0 occur with equal probability.

Split-Phase (Manchester Code):

In this signaling method, symbol 1 is represented by a positive half pulse of amplitude +A followed by a
negative half pulse of amplitude – A. For symbol 0, polarities of these two pulses are reversed.
If symbol ‘1’ is transmitted,
+A⁡⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁄2 ⁡⁡⁡⁡(half⁡interval)
x(t) = {
−A⁡⁡⁡⁡⁡for⁡⁡⁡ Tb ⁄2 ≤ t ≤ Tb ⁡⁡(half⁡interval)
If symbol ‘0’ is transmitted,
−A⁡⁡⁡⁡for⁡⁡⁡0 ≤ t ≤ Tb ⁄2 ⁡⁡⁡⁡(half⁡interval)
x(t) = {
+A⁡⁡⁡⁡⁡for⁡⁡⁡ Tb ⁄2 ≤ t ≤ Tb ⁡⁡(half⁡interval)
The unique property of this code is, it suppresses the DC component and has relatively insignificant low-
frequency components.

Fig: Line codes for the electrical representations of binary data

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 16

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

1. What is line coding? For the binary stream 011010 sketch the following line codes:
i) Unipolar NRZ
ii) Polar NRZ
iii) Unipolar RZ
iv) Bipolar RZ
v) Manchester June/July.2018, 15EC61, 06 Marks
Solution:

Power Spectra of Line Codes:

Unipolar NRZ Signaling: In this format the signal is
uniplor, its amplitude can be +A or zero. Hence, the signal
has some DC and most of the power lies between DC and
bit rate (1⁄T ) of the input signal. The power spectra is sinc
b
shaped and its main lobe extends from DC to (1⁄T ).
b
Power contained in frequencies above bit rate is very small.

Fig: Power Spectra

Unipolar RZ Signaling: In this format the signal is uniplor, its amplitude can be +A or zero. Hence, the
signal has some DC and most of the power lies between DC and bit rate (1⁄T ) of the input signal. The
b
power spectra is sinc shaped and its main lobe extends from DC to (1⁄T ). Power contained in frequencies
b

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 17

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

above bit rate is very small. An attractive feature is the presence of delta function at 𝑓 = 0, ± 1⁄T in the
b
power spectrum of the transmitted signal. The delta function can be used for bit-timing recovery at the
receiver.

Polar NRZ Signaling: In this format, the waveform takes positive as well as negative amplitudes. If
occurrence of binary ‘1’ or ‘0’ is not equal, then waveform has some DC value. The power spectra is sinc
pulse and contains most of the power from DC to bit rate (1⁄T ). The main lobe of sinc function is from
b
DC to bit rate frequency. Power contained in frequencies above bit rate is very small.

Fig: Power Spectra of Unipolar RZ Signaling Fig: Power Spectra of Polar NRZ Signaling

Bipolar RZ Signaling: In this format, successive 1’s are assigned pulses of alternating amplitudes. Hence,
waveform does not contain any DC component. The spectra is a pulse having peak power near 1⁄2T i.e.,
b
1
half bit rate. Thus power lies inside the bandwidth equal to bit rate ⁄T . The power content in frequencies
b
above bit rate is very small.

Split-Phase or Manchester Format: In this format, every symbol is transmitted with positive as well as
negative amplitude. Hence, there is no possibility of DC component in the signal. Most of the power lies in
the bandwidth of twice of bit rate (2⁄T ). Negligible power is contained at DC and (2⁄T ). Peak of the
b b
spectra occur somewhere near bit rate. The width of the main pulse is twice of other formats.

Fig: Power Spectra of Bipolar RZ Signaling Fig: Power Spectra of Manchester Signaling

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 18

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

Fig: Power Spectra of Various PAM Signals

Comparison of Line Codes:
1. Compare the power spectra of various line codes in terms of bandwidth, DC component, Noise
immunity and synchronization capability, with neat sketch.
Dec.2018/Jan.2019, 15EC61, 05 Marks
Sl. Unipolar Unipolar
Parameter Polar NRZ Bipolar RZ Manchester
No RZ NRZ
Nature/Polarities of Two levels Two levels Two levels Three levels Two levels
1
the signal 0 & +ve 0 & +ve –ve & +ve –ve, +ve & 0 –ve & +ve
DC component in May be
2 Present Present Absent Absent
the signal Present
Signal Frequency fb fb fb
3 fb fb
(Bandwidth) 2 2 2
Peak Power
4 High High Low High Low
requirement
5 Noise Immunity Poor Fair Fair Good Good
6 Cross Talk High High High Low Low
Synchronization
7 capability (effect of Better Poor Better Good Good
strings of 0’s & 1’s)

High Density Bipolar (HDB3) Signaling:

1. Write a note on HDB3 signaling. June/July.2018, 15EC61, 04 Marks
 In bipolar NRZ, there is no signal transmitted during the period when binary ‘0’ is present. This
creates a problem in synchronization when a long sequence of 0’s is present. This problem is
eliminated by adding pulses, when number of consecutive 0’s exceeds n.
 This type of coding is called as High Density Bipolar (HDB) coding denoted by HBDN. Here, N
can be 1, 2, 3… when N = 3, the code becomes HDB3, which is most widely used format.
 In the message sequence, when N + 1 zeros occurs, this group of zeros is replaced by special N + 1
binary digit sequences. These sequences contain some binary 1’s so that they can be detected at the
receiver. For HDB3 coding, those special sequences used are 000V and B00V.
 The below illustrates this type of coding. When there are four or more consecutive 0’s, the special
sequences 000V and B00V are inserted. Both B & V are considered as 1’s. Fig (c) shows the
representation of bipolar NRZ representation of this HDB3 coded message sequence.

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 19

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

Fig: (a) Message bits, (b) Coded sequence according to HDB3, (c) Bipolar NRZ format

Binary with N Zero Substitution (BNZS) Signaling:

This coding is used in DS-1 signals of the digital telephone hierarchy.

Binary with 8 Zero Substitution (B8ZS) Signaling:

 This coding replaces any string of 8 zeros in length with a sequence of 1’s and 0’s containing two
bipolar violations. For example, the sequence used is 000VB0VB. Here B = 1 confirms to bipolar
rule and V = 1 violates the bipolar rule. Therefore, V and B occurs in succession.
 The below figure shows the encoding of a binary sequence using B8ZS coding. The sequence of 8
zeros is replaced with 000VB0VB. Since B = 1 and V = 1, the replaced sequence will be 00011011.
Hence the new sequence is 01101 00011011 0110. This new sequence is encoded using bipolar
NRZ.

Fig: B8ZS encoding

Binary with 3 Zero Substitution (B3ZS) Signaling:

This coding is used in DS-3 signals of digital telephone hierarchy. This code uses B0V or 00V to replace
consecutive three zeros in the sequence. The below figure shows the waveforms of B3ZS coding.

Fig: B3ZS encoding

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 20

 Regulation – 2017 Digital Communication -17EC61

Module 1: Band-pass Signal to Equivalent Low-pass

Binary with 6 Zero Substitution (B6ZS) Signaling:
This coding is used in DS-2 signals of digital telephone hierarchy. It replaces the string of 6 zeros with
0VB0VB. The below figure, shows the waveforms of B3ZS coding.

Fig: B6ZS encoding

1. What is the advantage of HDB3 code over conventional alternate mark inversion (AMI) code. Code
the pattern “1010000011000011000000” using HDB3 encoding and AMI encoding.
Dec.2018/Jan.2019, 15EC61, 04 Marks

Prepared by: Halesha H R Sri Sairam College of Engineering Anekal. Page | 21