Chapter 3

Digital Modulation Techniques

Debre tabor University

Gafat Institute of Technology
Electrical and Computer Engineering

1
Overview
• Introduction to modulation and demodulation
• Basics of digital modulation
• Types of digital modulation techniques
• Digital modulation techniques in the Presence of Noise
• Optimum receiver for digital modulation techniques in the
presence of noise
• Multipath propagation mechanism and Mitigation
techniques

2
Introduction to modulation and demodulation
• Topics to be covered

3
Discussions
• What is modulation?
• What are the benefits of modulation?
• What are the types modulation?

4
 Modulation & demodulation
• Modulation is the process of encoding information from a
message source in a manner suitable for transmission
• It maps the coded information sequence into signal waveforms
that can be transmitted over the channel
• It could be done by varying either amplitude, phase,
frequency,(combination forms) of a carrier signal in
accordance with the message signal.
• It translates base-band message signal to band-pass signal
▪ Analogue and digital modulation
• Demodulation: Process of extracting the baseband message
▪ Coherent or non-coherent
5
Cont. …

6
 Some benefits of modulation
➢ Reduction of antenna height
𝜆 𝑐
▪ The size of antenna is ~ = 𝑤ℎ𝑒𝑟𝑒, 𝜆 is wave length of the RF wave.
4 4𝑓
▪ 3 KHz 25 Km antenna
▪ 3 GHz 2.5 cm antenna
➢ Avoids mixing of signals – translate signals to different EM spectrum
➢ Increase the range of communication
▪ Similar to throwing a paper by wrapping it around a stone
➢ Allows multiplexing
▪ Simultaneous transmission of several baseband signals over similar channel
➢ Bandwidth efficiency
▪ Effective use of limited frequency resources
➢ Improves quality of reception

7
Overview
• Basics of digital modulation
▪ Advantages of digital modulations
▪ Bit rate and Baud rate
▪ Performance measures
▪ Geometric representation and constellation diagrams
• Types of digital modulation techniques
• Digital modulation techniques in the Presence of Noise
• Optimum receiver for digital modulation techniques in the
presence of noise
• Multipath propagation mechanism and Mitigation techniques

8
 Analog vs Digital modulation
Advantages of digital modulation over its analog counterpart
• Greater noise immunity and robustness to channel
impairments
▪ Perform well in multipath and fading conditions
• Allow signal conditioning and processing techniques
▪ Error-control codes: Detect and/or correct transmission
errors
▪ Equalization: to improve the performance of the overall link
▪ Source coding
▪ Encryption: greater security
• Possible to be implemented completely in software
• Easier signal generation and processing

9
 Bit rate and Baud rate
• Bit rate: is the number of bits (data) that can be carried per
second
• Baud rate: is the number of signal units (or symbols) per
second used for carrying the bits (and achieve the Bit Rate).
▪ Baud rate can be less than or equal to the bit rate
• Note that each symbol can carry one or more bits
• Baud Rate = Bit rate / Number of bits per symbol
▪ The number of bits that can be carried by one symbol,
depends on the modulation technique used.
• Example: A modulated signal has a bit rate of 8000bps and
a baud rate of 1000 baud. How many bits are carried by each
signal element?
Bits/Baud = 8000/1000 = 8
10
 Performance measures
Two key performance measures of a modulation scheme are;
• Power efficiency: Describes the ability of a modulation
technique to preserve the fidelity of the digital message at low
power levels
• Fidelity: an acceptable bit error probability
• Often expressed as the ratio of the signal energy per bit (𝐸𝑏 ) to
the noise PSD (𝑁0 ) required to achieve a given probability of
error (say 10−5 ):

• is a measure of how favorably the tradeoff between fidelity and

signal power is made
11
 Cont. …
• Bandwidth efficiency: the ability of a modulation scheme to
accommodate data within a limited bandwidth
• It is defined as the ratio of the data rate R to the required RF
bandwidth B:

• Channel capacity gives an upper bound of achievable

bandwidth efficiency:

• It reflects how efficiently the allocated bandwidth is utilized

12
 Geometric representation
• Digital modulation involves choosing a particular analog signal
waveform si (t) from a finite set of S possible signal waveforms
based on the information bits applied to the modulator.
• For binary modulation schemes, a binary information bit is
mapped directly to a signal and S contains only 2 signals,
representing 0 and 1.
• For M-ary modulations, S contains more than 2 signals and
each represents more than a single bit of information. With a
signal set of size M, it is possible to transmit up to log 2 𝑀 bits
per signal.
• Any element of set S, 𝑆 = {s1 t , s2 (t), … , sM (t)} , can be
represented as a point in a vector space whose coordinates
are basis signals ∅j 𝑡 , 𝑗=1,2,…,N, such that N ≤ 𝑀 and

13
Cont. …
∞
න ∅𝑖 𝑡 ∅j 𝑡 𝑑𝑡 = 0, 𝑖 ≠ 𝑗; (Orthogonal)
−∞
∞
2
න ∅i 𝑡 𝑑𝑡 = 1, (𝑛𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑒𝑑)
−∞
• 𝑠𝑖 𝑡 can be represented as a linear combination of the
basis signals
𝑁

𝑠𝑖 𝑡 = ෍ 𝑠i𝑗 ∅j 𝑡 , 𝑖 = 1, 2, … , 𝑀
𝑗=0
Where,
𝑇
𝑠𝑖𝑗 = න 𝑠𝑖 𝑡 ∅𝑗 𝑡 𝑑𝑡
0

14
Cont. …

15
 Constellation diagram
• A graphical representation of the complex envelope of each
possible signal
• The 𝑥 −axis represents the in-phase component and the y-axis
represents the quadrature component of the complex envelope
• The distance between signals on a constellation diagram
relates to how different the modulation waveforms are and how
well a receiver can differentiate between them when random
noise is present.
• Examples

16
 Gram-Schmidt Orthogonalization procedure
• Suppose we have a set of M energy levels denoted by
𝑠1 𝑡 , 𝑠2 𝑡 , … , 𝑠𝑀 𝑡
• Starting with 𝑠1 𝑡 chosen from the list arbitrarily, the first basis
function is defined by
𝑠1 𝑡
∅1 𝑡 =
𝐸1
Where 𝐸1 is the energy of the signal 𝑠1 𝑡 . Then, we have
𝑠1 𝑡 = 𝐸1 ∅1 𝑡 = 𝑠11 ∅1 𝑡
• Next using the signal 𝑠2 𝑡 , we define the coefficient 𝑠21 as
𝑇
𝑠21 = න 𝑠2 𝑡 ∅1 𝑡
0
• We may thus introduce a new intermediate function

17
 Cont. …
𝑔2 𝑡 = 𝑠2 𝑡 − 𝑠21 ∅1 𝑡
• Now, the second basis function becomes;
𝑔2 𝑡 𝑠2 𝑡 − 𝑠21 ∅1 𝑡
∅2 𝑡 = =
𝑇 2 𝐸2 − 𝑠 2 21
‫׬‬0 𝑔2 (𝑡) 𝑑𝑡
• Continuing in this fashion, we may in general define
𝑖−1
𝑔𝑖 𝑡 = 𝑠𝑖 𝑡 − ෍ 𝑠𝑖𝑗 ∅𝑗 𝑡
𝑗=1
Where the coefficients themselves are defined by
𝑇
𝑠𝑖𝑗 = න 𝑠𝑖 𝑡 ∅𝑗 𝑡 𝑑𝑡, 𝑗 = 1, 2, … , 𝑖 − 1
0
• Given the 𝑔𝑖 𝑡 , we may now define the set of basis
functions as;
18
 Example- Gram Schmidt Orthogonalization
𝑔𝑖 𝑡
∅𝑖 𝑡 = , i = 1, 2, … , N
𝑇 2
‫׬‬0 𝑔𝑖 (𝑡) 𝑑𝑡
• Gram-Schmidt orthogonalization example ;
• Using Gram-Schmidt orthogonalization procedure, find a set of
orthonormal basis functions to represent the following 3 signals

19
Types of digital modulation techniques
• Digital modulation techniques can be classified in to the
following broad categories

20
Overview
• Introduction to modulation and demodulation
• Types of digital modulation techniques
▪ Binary modulation techniques
▪ M-ary modulation techniques
• Digital modulation techniques in the Presence of Noise
• Optimum receiver for digital modulation techniques in the
presence of noise
• Multipath propagation mechanism and Mitigation
techniques

21
 Binary modulation
• In binary modulation, the modulator produces one of two
distinct signals in response to one bit of source data at a time.

• In all cases,
▪ One bit is sent per symbol
▪ Baud rate = Bit rate
22
 Binary Amplitude Shift Keying (ASK)
• Amplitude of a carrier signal is varied to represent binary 1 or
0, while both frequency & phase remain constant
• Commonly, one of the amplitudes is zero

• The average energy per bit

𝐸+0
𝐸𝑏 = 𝑖. 𝑒 𝐸 = 2𝐸𝑏
2
• Decision region

23
 Cont. …
• ASK modulation wave form example

• Demodulation: only the presence or absence of a sinusoid in

a given time interval needs to be determined
• Advantage: simplicity
• Disadvantage: ASK is susceptible to noise and interference
• Application: is used to transmit digital data over optical fiber

24
 Binary Frequency Shift keying (BFSK)
• Frequency of carrier signal is varied to represent binary 1 or 0
while amplitude & phase remain constant during each bit
interval

• 𝑓1 and 𝑓2 are selected such that 𝑠1 𝑡 and 𝑠2 𝑡 are orthogonal.

• Two orthogonal basis functions are required

25
 Cont. ..
• Signal space representation and decision regions

• Demodulation: demodulator must be able to determine which

of two possible frequencies is present at a given time
• Advantage: FSK is less susceptible to errors than ASK
• Disadvantage: FSK is spectrum inefficient
• Application: over voice lines, in HF radio transmission.

26
 Cont. …
• FSK modulation waveform example

27
 Binary Phase shift Keying- (BPSK)
• Phase of carrier signal is varied to represent binary 1 or 0
while amplitude & freq. remain constant during each bit interval

• There is only one basis function

2
∅1 𝑡 = cos 2𝜋𝑓𝑐 𝑡 with 0 ≤ 𝑡 < 𝑇𝑏
𝑇𝑏

• Then 𝑠1 (𝑡) = 𝐸𝑏 ∅1 𝑡 and 𝑠2 𝑡 = − 𝐸𝑏 ∅1 𝑡

28
 Cont. …
• A BPSK system is characterized by a signal space that is one-
dimensional (i.e. N=1), and has two message points (i.e. M =2)
• Signal space representation and decision regions are;

• Demodulation: demodulator must determine the phase of

received sinusoid with respect to some reference phase
• Advantage: PSK is less susceptible to errors than ASK and
has more efficient bandwidth utilization compared to FSK
29
 Cont. …
• Disadvantage: more complex signal detection / recovery
process, than ASK and FSK
• PSK modulation wave form example

30
Overview
• Introduction to modulation and demodulation
• Types of digital modulation techniques
▪ Binary modulation techniques
▪ M-ary modulation techniques
• Digital modulation techniques in the Presence of Noise
• Optimum receiver for digital modulation techniques in the
presence of noise
• Mitigation techniques

31
 M-ary modulation
• In binary modulation, we send only one of two possible
signals during each bit interval 𝑇𝑏
• In M-ary modulation, we can send one of 𝑀 possible signals
during each signaling interval 𝑇
• In almost all applications, 𝑀 = 2𝑛 and 𝑇 = 𝑛𝑇𝑏 , where 𝑛 is
an integer
• Each of the M signals is called a symbol
• These signals are generated by changing the amplitude,
phase, frequency, or combined forms of a carrier in 𝑀
discrete steps.
• Thus, we have:

32
 M-ASK
• Amplitude of a carrier takes M different values
• For example, if M = 4, the ASK refers to four different
Amplitudes in which the carrier is sent.
• As 4 states are possible, two bits can be encoded per symbol
• In general, if number of possible states 𝑀 > 2, each symbol
can carry log 2 𝑀 bits.
𝑀 = 4, log 2 4 = 2 → 2 bitsൗsymbol
𝐴1 cos 2𝜋𝑓𝑐 (𝑡) 𝐴1 = 0: 00
𝐴2 cos 2𝜋𝑓𝑐 (𝑡) 𝐴2 = 1: 01
𝑆 𝑡 =
𝐴3 cos 2𝜋𝑓𝑐 (𝑡) 𝐴3 = 2: 10
𝐴4 cos 2𝜋𝑓𝑐 (𝑡) 𝐴4 = 3: 11

• This scheme is therefore more bandwidth efficient than BASK.

33
Cont. …
• Constellation plots of M-ASK

34
 M-FSK
• The frequency of the carrier takes on M possible values:
• In M-ary FSK modulation, the transmitted signals are defined
by
2𝐸𝑠 𝜋
𝑠𝑖 𝑡 = cos 𝑛𝑐 + 𝑖 𝑡 0 ≤ 𝑡 ≤ 𝑇𝑠 𝑖 = 1, 2, … , 𝑀
𝑇𝑠 𝑇𝑠
Where 𝑓𝑐 = 𝑛𝑐 Τ2𝑇𝑐 , for some fixed integer 𝑛𝑐 .
• The following diagram shows an example of 4-FSK waveform

𝐴 cos 2𝜋𝑓1 (𝑡) 00

𝐴 cos 2𝜋𝑓2 (𝑡) 01
𝑆 𝑡 =
𝐴 cos 2𝜋𝑓3 (𝑡) 10
𝐴 cos 2𝜋𝑓4 (𝑡) 11

35
 Cont.…
• The M transmitted signal frequencies are separated by 𝟏Τ𝟐𝑻𝒔 ,
making the signals orthogonal to one another.
• The average probability of error is given by
𝐸𝑠
𝑝𝑒 ≤ 𝑀 − 1 𝑄 𝑖𝑓 𝑐𝑜ℎ𝑒𝑟𝑒𝑛𝑡 𝑑𝑒𝑡𝑒𝑐𝑡𝑖𝑜𝑛 𝑖𝑠 𝑢𝑠𝑒𝑑
𝑁0
𝑀−1 −𝐸𝑠
𝑝𝑒 ≤ 𝑒𝑥𝑝 𝑖𝑓 𝑛𝑜𝑛 − 𝑐𝑜ℎ𝑒𝑟𝑒𝑛𝑡 𝑑𝑒𝑡𝑒𝑐𝑡𝑖𝑜𝑛 𝑖𝑠 𝑢𝑠𝑒𝑑
2 𝑁0
• The channel bandwidth of M-FSK signal can be defined as
𝑅𝑏 (𝑀 + 3)
𝐵𝑊 = 𝑖𝑓 𝑐𝑜ℎ𝑒𝑟𝑒𝑛𝑡 𝑑𝑒𝑡𝑒𝑐𝑡𝑖𝑜𝑛 𝑖𝑠 𝑢𝑠𝑒𝑑
2 log 2 𝑀
𝑅𝑏 𝑀
𝐵𝑊 = 𝑖𝑓 𝑛𝑜𝑛 − 𝑐𝑜ℎ𝑒𝑟𝑒𝑛𝑡 𝑑𝑒𝑡𝑒𝑐𝑡𝑖𝑜𝑛 𝑖𝑠 𝑢𝑠𝑒𝑑
2 log 2 𝑀

36
 M-PSK
• The phase of the carrier takes on M possible values, namely,
𝜃𝑖 = 2(𝑖 − 1)𝜋Τ𝑀, where 𝑖 = 1, 2, … , 𝑀
• The modulated waveform can be expressed as
2𝐸𝑠
𝑠𝑖 𝑡 = cos 2𝜋𝑓𝑐 𝑡 + 2(𝑖 − 1)𝜋Τ𝑀 , 0 ≤ 𝑡 ≤ 𝑇𝑠 𝑖 = 1, 2, … , 𝑀
𝑇𝑠
Where 𝐸𝑠 = 𝐸𝑏 log 2 𝑀 and 𝑇𝑠 = 𝑇𝑏 log 2 𝑀
• In quadrature form this equation can be written as;
2𝐸𝑠
𝑠𝑖 𝑡 = cos 2(𝑖 − 1)𝜋Τ𝑀 cos 2𝜋𝑓𝑐 𝑡 −
𝑇𝑠

2𝐸𝑠
sin 2(𝑖 − 1)𝜋Τ𝑀 sin 2𝜋𝑓𝑐 𝑡
𝑇𝑠

37
 Cont. …
2
• By choosing orthogonal basis functions Φ1 𝑡 = cos 2𝜋𝑓𝑐 𝑡
𝑇𝑠

2
and Φ2 𝑡 = sin 2𝜋𝑓𝑐 𝑡 , defined over 0 ≤ 𝑡 ≤ 𝑇𝑠 , the M-ary
𝑇𝑠
PSK signal can be expressed as;
𝑆𝑀−𝑃𝑆𝐾 𝑡 = 𝐸𝑠 cos 2(𝑖 − 1)𝜋Τ𝑀 Φ1 𝑡 − 𝐸𝑠 sin 2(𝑖 − 1)𝜋Τ𝑀 Φ2 𝑡
• Average symbol error probability M-Ary PSK:
2𝐸𝑏 log2 𝑀 𝜋
𝑝𝑒 ≤ 2𝑄 sin
𝑁0 𝑀

• Average symbol error probability for differential M-Ary PSK:

4𝐸𝑠 𝜋
𝑝𝑒 ≤ 2𝑄 sin
𝑁0 𝑀

38
 Cont. …
• For the case of QPSK (4-PSK)
𝐴 cos 2𝜋𝑓𝑐 𝑡 + 𝜑1 𝜑1 = 00 𝜑1 = 450 00
𝐴 cos 2𝜋𝑓𝑐 𝑡 + 𝜑2 𝜑1 = 900 𝑂𝑟 𝜑1 = 1350 01
𝑆 𝑡 =
𝐴 cos 2𝜋𝑓𝑐 𝑡 + 𝜑3 𝜑3 = 1800 𝜑3 = 2250 10
𝐴 cos 2𝜋𝑓𝑐 𝑡 + 𝜑4 𝜑4 = 2700 𝜑4 = 3150 11
• Constellation plots of M-ary PSK

• Thus, As M increases, the bandwidth efficiency increases but

the waveform energy (i.e., the power used to send the symbol)
must be increased to keep the BER at a certain level.
39
 M-ary QAM
• The principle of M-QAM is to have X possible variations in
phase and Y possible variations of amplitude
• X • Y possible variations
• The general form of M-ary QAM can be defined by
2𝐸𝑚𝑖𝑛 2𝐸𝑚𝑖𝑛
𝑆𝑖 𝑡 = 𝑎𝑖 cos 2𝜋𝑓𝑐 𝑡 + 𝑏𝑖 sin 2𝜋𝑓𝑐 𝑡
𝑇𝑠 𝑇𝑠
• Where 𝐸𝑚𝑖𝑛 is the energy of the signal with the lowest
amplitude and {𝑎𝑖 , 𝑏𝑖 } are given as;
−𝐿 + 1, 𝐿 − 1 −𝐿 + 3, 𝐿 − 1 … 𝐿 − 1, 𝐿 − 1
−𝐿 + 1, 𝐿 − 3 −𝐿 + 3, 𝐿 − 3 … 𝐿 − 1, 𝐿 − 3
𝑎𝑖 , 𝑏𝑖 = ⋮
⋮ ⋮ ⋮
(−𝐿 + 1, −𝐿 + 1) (−𝐿 + 3, −𝐿 + 1) … (𝐿 − 1, − 𝐿 + 1)
Where 𝐿 = 𝑀

40
 M-QAM
• Example: 8-QAM
▪ 2 different Amplitudes (𝐴1 = 1; 𝐴2 = 2)
▪ 4 different Phases (0𝑜 , 90𝑜 , 180𝑜 , 270𝑜 )
▪ 3 bits per symbol
• We can have numerous possible variations of phase shifts and
amplitude shifts, however the number of phase shifts should
be selected to be greater than number of amplitude shifts.
• 16-QAM
▪ There are sixteen QAM symbols 4 bits per symbol
• Average probability of error for M-ary QAM using coherent
detection can be described as;
1 2𝐸𝑚𝑖𝑛
𝑝𝑒 = 4 1 − 𝑄
𝑀 𝑁0

41
QAM Constellation examples

42
Selection of modulation systems

43
Cont. ..
• In the design of a digital communication system, there is
always a tradeoff between bandwidth efficiency and power
efficiency
• Example:
▪ Adding error coding to a message,
• Decreases bandwidth efficiency, but reduces the
required power for a particular bit error rate
▪ On the other hand, higher level modulation schemes
(M-ary keying)
• increases our bandwidth efficiency , but increase the
required received power
▪ Cost and implementation complexity
▪ Robustness to channel impairments

44
Bandwidth and power efficiency comparisons

45
Overview
• Introduction to modulation and demodulation
• Types of digital modulation techniques
• Digital modulation techniques in the Presence of Noise
• Optimum receiver for digital modulation techniques in the
presence of noise
• Multipath propagation mechanism and Mitigation
techniques

46
 Noise in digital modulation systems
• Noise is any unwanted signal that tend to disturb the transmission and
processing of signals in communication systems and over which we have
incomplete control.
• External sources: e.g., atmospheric, galactic noise, man-made-
interferences;
• Internal sources: generated by communication devices themselves.
▪ This type of noise represents a basic limitation on the performance of
electronic communication systems.
▪ Shot noise: arises in electronic devices such as diodes and transistors
because of the discrete nature of current flow in these devices.
▪ Thermal noise: caused by the rapid and random motion of electrons
within a conductor due to thermal agitation.

47
 Additive White Gaussian Noise (AWGN)
• The additive noise
• n(t) models all types of noise
• zero mean
• White noise
• Its power spectrum density (PSD) is constant over all frequencies, i.e.,

• Factor 1/2 is included to indicate that half the power is associated with
positive frequencies and half with negative.
• The term white is analogous to white light which contains equal
amounts of all frequencies (within the visible band of EM wave).
• Gaussian noise
• Represented in Gaussian random variable with zero mean and
variance of 𝜎𝑛 .

48
 Performances of modulation techniques in AWGN channel

• The channel is assumed to corrupt the signal by the

additive white Gaussian noise.
• Creates distortions in constellation

49
Performances of modulation techniques in AWGN channel

• ASK: symbol error probability

50
 Cont. …
• PSK: symbol error probability

51
 Cont. …
• QAM: symbol error probability

52
Cont. …
• FSK: symbol error probability

53
Overview
• Introduction to modulation and demodulation
• Types of digital modulation techniques
• Digital modulation techniques in the Presence of Noise
• Optimum receiver for digital modulation techniques in the
presence of noise
• Multipath propagation mechanism and Mitigation techniques

54
Optimum receivers in the presence of noise
• A channel model for the received signal over an AWGN
channel is depicted below

• n(t) is the sample function of the AWGN process with

power spectral density

55
 Optimum Receivers
• Objective: Based on the observation of a received signal r(t)
over the signal interval, design a receiver that is optimum in
the sense that it minimizes the probability of error
• Consider the following receiver configuration

• Signal demodulation: Converts the received waveform into an

N-dimensional vector
𝐫 = 𝐫1 , 𝐫2 , 𝐫3 , … , 𝐫𝑁
• Detector: Decides which of the M possible signal waveforms
are transmitted based on the vector r

56
Overview
• Introduction to modulation and demodulation
• Types of digital modulation techniques
• Digital modulation techniques in the Presence of Noise
• Optimum receiver for digital modulation techniques in the
presence of noise
▪ Correlation demodulator
▪ Matched filter demodulator
▪ Optimal detector
• Multipath propagation mechanism and Mitigation techniques

57
 Correlation Demodulator
• Correlation demodulator: Decomposes 𝑟(𝑡) into N dimensional
vectors
• The signal and noise are expanded into a series of linearly
weighted orthonormal basis functions {∅N 𝑡 } which spans the
signal space so that all possible members of the signal set
{𝑠𝑚 (𝑡), 1 ≤ 𝑚 ≤ 𝑀} can be represented
• The basis functions do not span the noise space
• However, the noise terms that fall outside the signal space
are irrelevant to the detection of the required signal

58
Correlation type demodulator

59
 Cont. …
• The N correlators essentially compute the projection of r(t)
onto the N basis functions {∅𝑵 (𝒕)} such that
𝑇 𝑇

න 𝑟(𝑡)∅𝑘 (𝑡) 𝑑𝑡 = න 𝑠𝑚 𝑡 + 𝑛 𝑡 ∅𝑘 𝑡 𝑑𝑡
0 0
𝑟𝑘 = 𝑠𝑚𝑘 + 𝑛𝑘 , 𝑘 = 1, 2, … , 𝑁, where,
𝑇

𝑠𝑚𝑘 = න 𝑠𝑚 𝑡 ∅𝑘 𝑡 𝑑𝑡 , 𝑘 = 1, 2, … , 𝑁
0
𝑇

𝑛𝑘 = න 𝑛 𝑡 ∅𝑘 𝑡 𝑑𝑡 , 𝑘 = 1, 2, … , 𝑁
0
• The signal is now represented by the vector 𝒔mk = [𝒔m1 , 𝒔m2 , … , 𝒔m𝑁 ]
whose values depend on which of the M possible signals was transmitted

60
 Cont. …
• The components {𝑛𝑘 } are random variables that arise from the
additive white Gaussian noise
• In the interval 0 ≤ t ≤ T
𝑁 𝑁 𝑁

𝑟 𝑡 = ෍ 𝑠𝑚𝑘 ∅𝑘 𝑡 + ෍ 𝑛𝑘 𝑡 ∅𝑘 𝑡 + 𝑛′ (𝑡) = ෍ 𝑟𝑘 ∅𝑘 𝑡 + 𝑛′ (𝑡)

𝑘=1 𝑘=1 𝑘=1
𝑁

𝑤ℎ𝑒𝑟𝑒, 𝑛′ 𝑡 = 𝑛 𝑡 − ෍ 𝑛𝑘 𝑡 ∅𝑘 𝑡
𝑘=1
• 𝑛′ 𝑡 is a zero-mean Gaussian noise process that represents
the difference between original noise process 𝑛(𝑡) and part
corresponding to projection of 𝑛(𝑡) onto {∅𝑘 𝑡 }

61
 Cont. …
• Recall that 𝑛’(𝑡) is irrelevant to the decision as to which signal
is transmitted and thus the decision can be based on the
correlator output signal and noise components
i.e., 𝑟𝑘 = 𝑠𝑚𝑘 + 𝑛𝑘 , 𝑘 = 1, 2, … , 𝑁
• Note that the signal components are deterministic and the
noise components are Gaussian with zero-mean
• Thus, {𝑛𝑘 } are zero-mean, uncorrelated Gaussian random
variables (also independent) with common variance

• Further, {𝑟𝑘 } is also Gaussian with mean 𝑠𝑚𝑘 and the same
variance

62
 Cont. …
• The output { 𝑟𝑘 } conditioned on the 𝑚𝑡ℎ signal being
transmitted are also statistically independent random
variables with probability density function given by

• So that the joint conditional PDF is

63
Overview
• Introduction to modulation and demodulation
• Types of digital modulation techniques
• Digital modulation techniques in the Presence of Noise
• Optimum receiver for digital modulation techniques in the
presence of noise
▪ Correlation demodulator
▪ Matched filter demodulator
▪ Optimal detector
• Multipath propagation mechanism and Mitigation techniques

64
 Matched filter Demodulator
• Here, we employ a bank of N-linear filters instead of using N
correlators to generate {𝑟𝑘 }
• The impulse response of these filters is matched to the
basis functions such that
ℎ𝑘 𝑡 = ∅𝑘 𝑇 − 𝑡 , 0≤𝑡≤𝑇
• The output of these filters are given by
𝑡 𝑡

𝑦𝑘 𝑡 = න 𝑟(𝜏) ℎ𝑘 𝑡 − 𝜏 𝑑𝜏 = න 𝑟(𝜏) ∅𝑘 𝑇 − 𝑡 + 𝜏 𝑑𝜏, 𝑘 = 1,2, … , 𝑁

0 0
• Sampling the output of the filters at t = T, we obtain
𝑇

𝑦𝑘 𝑇 = න 𝑟(𝜏) ∅𝑘 𝜏 𝑑𝜏 = 𝑟𝑘 , 𝑘 = 1,2, … , 𝑁
0

65
 Cont. …
• Hence, the sample outputs of the filter at time t = T are the set
of values {𝑟𝑘 } obtained from the N linear correlators
• In general, for a signal 𝑆(𝑡) and matched filter ℎ(𝑡) = 𝑆(𝑇 − 𝑡)

66
Cont. …

67
Cont. …
• Example: The M=4 biorthogonal signals have dimension
N=2; and we need two basis functions to represent them
• Choose the two basis functions such that

• The impulse response of the two matched filters are;

68
Overview
• Introduction to modulation and demodulation
• Types of digital modulation techniques
• Digital modulation techniques in the Presence of Noise
• Optimum receiver for digital modulation techniques in the
presence of noise
▪ Correlation demodulator
▪ Matched filter demodulator
▪ Optimal detector
• Multipath propagation mechanism and Mitigation techniques

69
 Optimal detectors
• Output of the correlator or the matched filter demodulator
produces the vector 𝐫 = 𝐫1 , 𝐫2 , 𝐫3 , … , 𝐫𝑁
• This output vector contains all the relevant information about
the received signal waveform
• On the basis of the vector r, we need to make a decision on
what is transmitted such that the decision is optimal
• Optimality criterion: Decide on what is transmitted in each
interval based on the observation of the vector r in each
interval such that the probability of correct decision is
maximized (or the probability of error is minimized)
• Define the a posteriori probability as
P{signal 𝑠𝑚 was transmitted} = ; m = 1,2,…M

70
 Cont. …
• Decision criterion: Select the signal corresponding to the
maximum of the set of a posteriori probabilities
▪ The criterion maximizes the probability of correct decision and,
hence, minimizes the probability of error
• This decision criterion is called maximum a posteriori
probability (MAP) criterion
• Note that:

• P{𝒔𝒎 } is the apriori probability of the 𝑚𝑡ℎ signal being

transmitted
• Note also that:

71
 Cont. …

• The computation of the a posteriori probability 𝑃(𝒔𝑚 │𝑟)

requires knowledge of the:
▪ Apriori probability 𝑃(𝒔𝑚 ) and
▪ Conditional pdf 𝑃(𝒓|𝒔𝑚 )
• For the special case where M signals are equi-probable such
that the apriori probability 𝑃(𝒔𝑚 ) = 1/𝑀 for all M
• Further 𝑃(𝒓) is independent of which signal is transmitted
• Then maximizing 𝑃(𝒔𝑚 │𝒓) is equivalent to finding the
signal that maximizes P(r|𝒔𝑚 )
• The conditional pdf 𝑃(𝒓|𝒔𝑚 ), or any monotonic function of it,
is called the likelihood function
• It is referred to as the maximum likelihood (ML) criterion

72
 Cont. …
• Note that the MAP and ML criteria are equivalent when the
apriori probabilities 𝑃(𝒔𝑚 ) are all equal
▪ I.e {𝒔𝑚 } are equiprobable
• For the AWGN channel, the likelihood function becomes

• Taking the logarithms of both sides simplifies the computation

• Finding the maximum of ln(𝑃(𝒓|𝒔𝑚 )) over 𝒔𝑚 is equivalent to

finding the signal 𝒔𝑚 that minimizes the Euclidean distance

73
 Cont. …
• The decision rule on the ML criterion is equivalent to finding
the signal 𝒔𝑚 that is closest in distance to the received signal
vector r
▪ Called the minimum distance detection
• Expanding the distance metrics

• Since ||r||2 is common to all the decision metrics, and hence

it may be ignored in the computation of the metrics
• Then the modified distance metrics becomes;

74
 Cont. …
• Note: 𝒓. 𝒔𝑚 represents the projection of the received signal
vector onto each of the M possible transmitted vectors
• Thus, it measures the correlation between the received vector
and the 𝑚𝑡ℎ signal
• Note also that ||sm ||2 = 𝜀𝑠𝑚 can be thought of as a bias term
that serves as compensation for signal sets that have unequal
energies such as ASK
• If all signals have equal energy, it may be ignored in the
computation of the correlation metrics and the distance
metrics D and D’
• The correlation metrics can thus be computed as

75
 Cont. …
• These metrics can be generated by a demodulator that
▪ Cross-correlates the received signal r(t) with each M
possible transmitted signal and
▪ Adjust each correlator output for the bias
▪ Select the signal corresponding to the largest correlation
metrics

76
 Cont. …
• Note that for signals of unequal energies the output of the
correlators are adjusted by ½ 𝜀𝑠𝑚
• Alternatively, 𝑟(𝑡) could be passed through a bank of N
matched filters and sampled at 𝑡 = 𝑇
• For non-equiprobable signal we apply MAP based on the
probabilities 𝑃(𝒔𝑚 │𝒓); m = 1,2,…..M
• Or, alternatively on the metrics

77
 Optimum detector example
• Example: Consider a binary ASK with 𝑠1 = −𝑠2 = 𝜀𝑏 , where
𝜀𝑏 is the energy per bit
• Let the a priori probabilities be 𝑃(𝑠1 ) = 𝑝 and 𝑃(𝑠2 ) = 1 − 𝑝
• Question: Determine the metrics for an optimum MAP detector
when the transmitted signal is corrupted by AWGN?
• Solution: The received signal vector (one-dimensional) for
binary ASK is
𝑟 = ± 𝜀𝑏 + 𝑦𝑛 (𝑡)
• Where 𝑦𝑛 (𝑡) is a zero-mean Gaussian random variable with
2 1
variance 𝜎 = 𝑁0
2
• Consequently, the conditional PDFs of 𝑝 𝑟 𝑠𝑚 for the two
signals is

78
 Cont. …

• If 𝑃𝑀 𝑟, 𝑠1 > 𝑃𝑀 𝑟, 𝑠2 , then select 𝑠1 as the transmitted

signal otherwise select 𝑠2 .
• The decision rule may be expressed as

79
 Cont. …
• Which upon substitution for 𝑃𝑀 𝑟, 𝑠1 and 𝑃𝑀 𝑟, 𝑠2 gives

• Hence, the optimum detector computes the product r 𝜀𝑏 and

compares it with the threshold 𝜏ℎ
▪ If r 𝜀𝑏 > 𝜏ℎ the 𝑠1 is transmitted or
▪ If r 𝜀𝑏 < 𝜏ℎ then decides on 𝑠2
• Note that the threshold depends on 𝑝 and 𝑁0
1
▪ For 𝑝 = , 𝜏ℎ = 0
2
1
▪ For 𝑝 > ,𝜏 < 0, then 𝑠1 is more probable, or vice versa
2 ℎ

80
`

3/27/2023 81