Chapter 3:

BASEBAND PULSE AND DIGITAL SIGNALING

➢ Chapter Objectives:
• Analog-to-digital signaling (pulse code modulation )
Binary and multilevel digitals signals
• Spectra and bandwidths of digital signals
• Prevention of intersymbol interference
• Time division multiplexing
• Packet transmission

1
 INTRODUCTION
➢ This chapter we study how to encode analog waveforms into base
band digital signals. Digital signal is popular because of the low
cost and flexibility.
➢ Main goals:
• To study how analog waveforms can be converted to digital
waveforms, Pulse Code Modulation.
• To learn how to compute the spectrum for digital signals.
• Examine how the filtering of pulse signals affects our ability to
recover the digital information. Intersymbol interference (ISI).
• To study how we can multiplex (combine) data from several
digital bit streams into one high-speed digital stream for
transmission over a digital system Time-division Multiplexing.

2
 PULSE AMPLITUDE MODULATION
➢ Pulse Amplitude Modulation (PAM) is used to describe the
conversion of the analog signal to a pulse-type signal in which the
amplitude of the pulse denotes the analog information.

➢ The purpose of PAM signaling is to provide another waveform that

looks like pulses, yet contains the information that was present in
the analog waveform.

➢ There are two classes of PAM signals:

• PAM that uses Natural Sampling (gating);
• PAM that uses Instantaneous Sampling to produce a flat-top
pulse.

3
 Natural Sampling (Gating)

DEFINTION: If w(t) is an analog waveform bandlimited to B hertz, the PAM

signal that uses natural sampling (gating) is

ws(t) =w(t)s(t) Where

S(t) is a rectangular wave switching waveform and fs = 1/Ts ≥ 2B.

THEORM: The spectrum for a naturally sampled PAM signal is:
 
sin( nd )
Ws ( f ) = F[ws (t )] =  cnW ( f − nf s ) = d
n =−

n =−  nd
W ( f − nf s )

• Where fs= 1/Ts, ωs = 2π fs,

• the Duty Cycle of s(t) is d = τ/Ts ,
• W(f)= F[w(t)] is the spectrum of the original unsampled waveform,
• cn represents the Fourier series coefficients of the switching waveform.

4
 Natural Sampling (Gating)
w(t)

s(t)

ws(t) =w(t)s(t)

5
 Generating Natural Sampling
➢ The PAM wave form with natural sampling can be generated using a
CMOS circuit consisting of a clock and analog switch as shown.

6
 Spectrum of Natural Sampling
 
sin( nd )
Ws ( f ) = F[ws (t )] =  cnW ( f − nf s ) = d
n =−

n =−  nd
W ( f − nf s )

• The duty cycle of the switching

waveform is d = τ/Ts = 1/3.
• The sampling rate is fs = 4B.


sin( nd )
Ws ( f ) = d 
n =−  nd
W ( f − nf s )

sin( nd )
d
 nd

7
 Recovering Naturally Sampled PAM
➢ At the receiver, the original analog waveform, w(t), can be recovered
from the PAM signal, ws(t), by passing the PAM signal through a low-
pass filter where the cutoff frequency is: B <fcutoff < fs -B

➢ If the analog signal is under sampled fs < 2B, the effect of spectral
overlapping is called Aliasing. This results in a recovered analog
signal that is distorted compared to the original waveform.

LPF Filter
B <fcutoff < fs -B
8
 Demodulation of PAM Signal
➢ The analog waveform may be recovered from the PAM signal
by using product detection,

9
 Instantaneous Sampling (Flat-Top PAM)
• This type of PAM signal
consists of instantaneous
samples.

• w(t) is sampled at t = kTs .

• The sample values w(kTs )

determine the amplitude of
the flat-top rectangular
pulses.

10
 Instantaneous Sampling (Flat-Top PAM)
➢ DEFINITION: If w(t) is an analog waveform bandlimited to B
Hertz, the instantaneous sampled PAM signal is given by
 
 

ws (t ) =  w(kTs )h(t − kTs ) = h(t )   w(kTs ) (t − kTs ) = h(t )   w(t )   (t − kTs ) 
k =− k =−  k =− 
– Where h(t) denotes the sampling-pulse shape and, for flat-top sampling,
the pulse shape is,

THEOREM: The spectrum for a flat-top PAM signal is:


1
Ws ( f ) = H ( f )  W ( f − nf s )
Ts k =−

 sin  f 
H ( f ) =  h(t )  =   
  f 
11
 The spectrum of the flat-top PAM
➢ Analog signal maybe recovered from the flat-top PAM signal by the use of a LPF.
LPF Response
Note that the recovered signal
has some distortions due to the
curvature of the H(f).
Distortions can be removed by
using a LPF having a response
1/H(f).

12
 Some notes on PAM
• The flat-top PAM signal could be generated by using a sample-
and-hold type electronic circuit.
• There is some high frequency loss in the recovered analog
waveform due to filtering effect H(f) caused by the flat top pulse
shape.
• This can be compensated (Equalized) at the receiver by making the
transfer function of the LPF to 1/H(f)
• This is a very common practice called “EQUALIZATION”
• The pulse width τ is called the APERTURE since τ/Ts determines
the gain of the recovered analog signal

➢ Disadvantages of PAM
• PAM requires a very larger bandwidth than that of the original signal;
• The noise performance of the PAM system is not satisfying.

13
 Chapter 3:
Pulse Code Modulation
➢ Pulse Code Modulation
➢ Quantizing
➢ Encoding
➢ Analogue to Digital Conversion
➢ Bandwidth of PCM Signals

14
 PULSE CODE MODULATION (PCM)
➢ DEFINITION: Pulse code modulation (PCM) is essentially
analog-to-digital conversion of a special type where the
information contained in the instantaneous samples of an analog
signal is represented by digital words in a serial bit stream.

➢ The advantages of PCM are:

• Relatively inexpensive digital circuitry may be used extensively.
• PCM signals derived from all types of analog sources may be merged with
data signals and transmitted over a common high-speed digital
communication system.
• In long-distance digital telephone systems requiring repeaters, a clean PCM
waveform can be regenerated at the output of each repeater, where the input
consists of a noisy PCM waveform.
• The noise performance of a digital system can be superior to that of an
analog system.
• The probability of error for the system output can be reduced even further
by the use of appropriate coding techniques.

15
 Sampling, Quantizing, and Encoding
➢ The PCM signal is generated by carrying out three basic operations:
1. Sampling
2. Quantizing
3. Encoding
1. Sampling operation generates a flat-top PAM signal.
2. Quantizing operation approximates the analog values by using a
finite number of levels. This operation is considered in 3 steps
a) Uniform Quantizer
b) Quantization Error
c) Quantized PAM signal output
3. PCM signal is obtained from the quantized PAM signal by encoding
each quantized sample value into a digital word.

16
 Analog to Digital Conversion
➢ The Analog-to-digital Converter (ADC)
performs three functions:
Analog – Sampling
Input • Makes the signal discrete in time.
Signal
• If the analog input has a bandwidth
Sample
of W Hz, then the minimum sample
frequency such that the signal can be
reconstructed without distortion.
ADC – Quantization
Quantize • Makes the signal discrete in
111
110
101
amplitude.
100
011
• Round off to one of q discrete levels.
– Encode
010
001
Encode 000

• Maps the quantized values to digital

words that are  bits long.
➢ If the (Nyquist) Sampling Theorem is
Digital Output satisfied, then only quantization introduces
Signal distortion to the system.
111 111 001 010 011 111 011
17
 Quantization
➢ The output of a sampler is still continuous in amplitude.
– Each sample can take on any value e.g. 3.752, 0.001, etc.
– The number of possible values is infinite.

➢ To transmit as a digital signal we must restrict the number of

possible values.

➢ Quantization is the process of “rounding off” a sample according to

some rule.
– E.g. suppose we must round to the nearest tenth, then:
3.752 --> 3.8 0.001 --> 0

18
Illustration of the Quantization Error

19
 Uniform Quantization
Dynamic Range: • Most ADC’s use uniform
(-8, 8)
quantizers.
Output sample
• The quantization levels of a
7
XQ
5
uniform quantizer are
3

1
equally spaced apart.
-8 -6 -4 -2 -1 2 4 6 8
• Uniform quantizers are
Input sample X
-3
optimal when the input
-5 distribution is uniform.
-7 When all values within the
Quantization Characteristic Dynamic Range of the
Example: Uniform  =3 bit quantizer quantizer are equally likely.
q=8 and XQ = {1,3,5,7}

20
Quantization Example

Analogue signal

Sampling TIMING

Quantization levels.
Quantized to 5-levels

Quantization levels
Quantized 10-levels

21
 PCM encoding example

Levels are encoded

using this table

Table: Quantization levels with belonging code words

M=8

Chart 2. Process of restoring a signal.

Chart 1. Quantization and digitalization of a signal. PCM encoded signal in binary form:
Signal is quantized in 11 time points & 8 quantization segments. 101 111 110 001 010 100 111 100 011 010 101
Total of 33 bits were used to encode a signal

22
 Encoding
• The output of the quantizer is one of M possible signal levels.
– If we want to use a binary transmission system, then we need to map
each quantized sample into an n bit binary word.

M = 2n , n = log 2 ( M )
• Encoding is the process of representing each quantized sample
by an  bit code word.
– The mapping is one-to-one so there is no distortion introduced by
encoding.
– Some mappings are better than others.
• A Gray code gives the best end-to-end performance.
• The weakness of Gray codes is poor performance when the sign bit
(MSB) is received in error.

23
 Gray Codes
• With gray codes adjacent samples differ only in one bit position.
• Example (3 bit quantization):
XQ Natural coding Gray Coding
+7 111 110
+5 110 111
+3 101 101
+1 100 100
-1 011 000
-3 010 001
-5 001 011
-7 000 010
• With this gray code, a single bit error will result in an amplitude
error of only 2.
– Unless the MSB is in error.

24
 Waveforms in a PCM system for M=8
M=8

(a) Quantizer Input output characteristics

(b) Analog Signal, PAM Signal, Quantized PAM Signal

M = 2n n = log 2 ( M )
M is the number of Quantization levels
(c) Error Signal n is the number of bits per sample

(d) PCM Signal 25

PCM Transmission System

26
 Practical PCM Circuits
• Three popular techniques are used to implement the
analog-to-digital converter (ADC) encoding
operation:
1. The counting or ramp, ( Maxim ICL7126 ADC)
2. Serial or successive approximation, (AD 570)
3. Parallel or flash encoders. ( CA3318)
• The objective of these circuits is to generate the
PCM word.
• Parallel digital output obtained (from one of the
above techniques) needs to be serialized before
sending over a 2-wire channel
• This is accomplished by parallel-to-serial
converters [Serial Input-Output (SIO) chip]
• UART,USRT and USART are examples for SIO’s

27
 Bandwidth of PCM Signals
• The spectrum of the PCM signal is not directly related to the spectrum of the
input signal.
• The bandwidth of (serial) binary PCM waveforms depends on the bit rate R
and the waveform pulse shape used to represent the data.
• The Bit Rate R is
R=nfs

Where n is the number of bits in the PCM word (M=2n) and fs is the sampling
rate.
• For no aliasing case (fs≥ 2B), the MINIMUM Bandwidth of PCM Bpcm(Min) is:

Bpcm(Min) = R/2 = nfs//2

The Minimum Bandwidth of nfs//2 is obtained only when sin(x)/x pulse is used
to generate the PCM waveform.

• For PCM waveform generated by rectangular pulses, the First-null Bandwidth

is:

Bpcm = R = nfs
28
 Chapter 3:
PCM Noise and Companding
➢ Quantization Noise
➢ Signal to Noise Ratio
➢ PCM Telephone System
➢ Nonuniform Quantization
➢ Companding

29
 Quantization Noise
➢ The process of quantization can be interpreted as an additive noise
process.
Signal Quantized Signal
X XQ

Quantization Noise
nQ

• The signal to quantization noise ratio (SNR)Q=S/N is given as:

Average Power{ X }
( SNR)Q =
Average Power{nQ }

30
 Effects of Noise on PCM
➢ Two main effects produce the noise or distortion in the PCM output:
– Quantizing noise that is caused by the M-step quantizer at the PCM transmitter.
– Bit errors in the recovered PCM signal, caused by channel noise and improper filtering.

• If the input analog signal is band limited and sampled fast enough so that the
aliasing noise on the recovered signal is negligible, the ratio of the recovered analog
peak signal power to the total average noise power is:

• The ratio of the average signal power to the average noise power is

– M is the number of quantized levels used in the PCM system.

– Pe is the probability of bit error in the recovered binary PCM signal at the receiver DAC
before it is converted back into an analog signal.

31
 Effects of Quantizing Noise
• If Pe is negligible, there are no bit errors resulting from channel noise and no ISI, the
Peak SNR resulting from only quantizing error is:

• The Average SNR due to quantizing errors is:

• Above equations can be expresses in decibels as,

Where, M = 2n
α = 4.77 for peak SNR
α = 0 for average SNR

32
Determining number of bits at
the output of PCM and Signal
to Noise Ratio due to
𝑆
Quantization Error
𝑁 𝑄
 Quantization Error
As mentioned before, Quantization error
is the difference between the Analog
signal and the Quantized signal.
The number of bits which represent each
sample is directly related to the
Quantization Error.
Vpp
The relationship is given by:
Peak to peak
50 amplitude
𝑛 ≥ 3,32 log10
𝑃
P: PCM accuracy which is given as a
percentage. It is the ratio of peak to peak
quantization error to the peak to peak Quantization Error
amplitude of the analog signal.
Number of Quantization Levels: M=2n
Amplitude range for a level:
𝑉𝑃𝑃
∆=
𝑀
 Example
What should be the number of Quantization
levels for the following signal if the PCM
accuracy shall not exceed %0.5?

50
𝑛 ≥ 3,32 log10
𝑃
50 50
𝑛𝑚𝑖𝑛 = 3,32 log10 = 3,32 log10 = 6.64
𝑃 0,5

Since there should be an integer number of

bits, n should be taken as 7. (Rounded UP)

So, the number of Quantization Levels is: M=2n=27=128.

The amplitude range for a level is:
10
∆= =0.0781V=78.1mV
128
 Average SNR due to
𝑆
Quantization error
𝑁 𝑄

• Average SNR due to Quantization error:

𝑆
= 𝑀2
𝑁 𝑄
Exll:
For the previous example the Average SNR due to
Quantization error :
𝑆
= 1282 =16384
𝑁 𝑄
𝑆
= 10 log 16384 = 42.1𝑑𝐵
𝑁 𝑄 𝑑𝐵
 DESIGN OF A PCM SIGNAL FOR TELEPHONE SYSTEMS
• Assume that an analog audio voice-frequency(VF) telephone signal occupies a band from
300 to 3,400Hz. The signal is to be converted to a PCM signal for transmission over a
digital telephone system. The minimum sampling frequency is 2x3.4 = 6.8 ksample/sec.
• To be able to use of a low-cost low-pass antialiasing filter, the VF signal is oversampled
with a sampling frequency of 8ksamples/sec.
• This is the standard adopted by the Unites States telephone industry.
• Assume that each sample values is represented by 8 bits; then the bit rate of the binary
PCM signal is

• This 64-kbit/s signal is called a DS-0 signal (digital signal, type zero).
• The minimum absolute bandwidth of the binary PCM signal is

R nf s
BPCM  =
2 2
This B is for a sinx/x type pulse sampling

37
 DESIGN OF A PCM SIGNAL FOR TELEPHONE SYSTEMS

• If we use a rectangular pulse for sampling the first null bandwidth is given by

• We require a bandwidth of 64kHz to transmit this digital voice PCM signal, whereas the
bandwidth of the original analog voice signal was, at most, 4kHz.

• We observe that the peak signal-to-quantizing noise power ratio is:

Note:
1. Coding with parity bits does NOT affect the quantizing noise,
2. However coding with parity bits will improve errors caused by channel
or ISI, which will be included in Pe ( assumed to be 0).
38
PCM Example:
 Nonuniform Quantization
➢ Many signals such as speech have a nonuniform distribution.
– The amplitude is more likely to be close to zero than to be at higher levels.
➢ Nonuniform quantizers have unequally spaced levels
– The spacing can be chosen to optimize the SNR for a particular type of signal.
Output sample
XQ 6

2 Example: Nonuniform 3 bit quantizer

-8 -6 -4 -2 2 4 6 8

-2
Input sample
X
-4

-6

40
 Companding
• Nonuniform quantizers are difficult to make and expensive.
• An alternative is to first pass the speech signal through a
nonlinearity before quantizing with a uniform quantizer.
• The nonlinearity causes the signal amplitude to be
Compressed.
– The input to the quantizer will have a more uniform
distribution.
• At the receiver, the signal is Expanded by an inverse to the
nonlinearity.
• The process of compressing and expanding is called
Companding.

41
 -Law Companding

• Telephones in the U.S., Canada and

Japan use -law companding:
Output |x(t)|

ln(1 +  | x(t )|)

| y (t ) |=
ln(1 +  )
– Where  = 255 and |x(t)| < 1

0 1
Input |x(t)|

42
 Non Uniform quantizing
• Voice signals are more likely to have amplitudes near zero than at extreme peaks.
• For such signals with non-uniform amplitude distribution quantizing noise will be
higher for amplitude values near zero.
• A technique to increase amplitudes near zero is called Companding.

Effect of non linear quantizing can be

can be obtained by first passing the
analog signal through a compressor
and then through a uniform quantizer.

x x’ x’ y
Q(.)
C(.)
Compressor Uniform Quantizer

43
 Example: -law Companding
1

0.5

x[n]=speech /song/ 0

-0.5

-1
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

0.5

y[n]=C(x[n]) 0

-0.5
Companded Signal -1
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

0.5
Close View of the Signal
Segment of x[n] 0

-0.5

-1
2200 2300 2400 2500 2600 2700 2800 2900 3000

0.5
Segment of y[n] 0

Companded Signal -0.5

-1
2200 2300 2400 2500 2600 2700 2800 2900 3000

44
 A-law and −law Companding
• These two are standard companding methods.
• u-Law is used in North America and Japan
• A-Law is used elsewhere to compress digital telephone signals

45
 SNR of Compander
• The output SNR is a function of input signal level for uniform quantizing.
• But it is relatively insensitive for input level for a compander

46
 SNR Performance of Compander

• The output SNR is a function of input signal level for uniform quantizing.
• But it is relatively insensitive for input level for a compander.

• α = 4.77 - 20 Log ( V/xrms) for Uniform Quantizer

V is the peak signal level and xrms is the rms value

• α = 4.77 - 20 log[Ln(1 + μ)] for μ-law companding

• α = 4.77 - 20 log[1 + Ln A] for A-law companding

47
 V.90 56-Kbps PCM Computer modem
• The V.90 PC Modem transmits data at 56kb/s from a PC
via an analog signal on a dial-up telephone line.
• A μ law compander is used in quantization with a value
for μ of 255.
• The modem clock is synchronized to the 8-ksample/ sec
clock of the telephone company.
• 7 bits of the 8 bit PCM are used to get a data rate of
56kb/s ( Frequencies below 300Hz are omitted to get rid
of the power line noise in harmonics of 60Hz).
• SNR of the line should be at least 52dB to operate on
56kbps.
• If SNR is below 52dB the modem will fallback to lower
speeds ( 33.3 kbps, 28.8kbps or 24kbps).

48
 Digital Signaling

➢ Digital Signaling
➢ Vector Representation
➢ Bandwidth Estimation
➢ Binary Signaling
➢ Multilevel Signaling

49
 Digital Signaling
➢ Mathematical Representation of the waveform: Voltage (or current) waveform
for digital signals:
N
w ( t ) =  wkk ( t ) 0  t  T0
k =1

wk Digital Data, w ( t ) PCM word or message

k (t ) k=1, 2, 3 N N Orthogonal functions
N Number of dimensions required to describe w(t )
Message time span ~ T0 seconds
➢ Example: Message ‘X’ from a digital source - code word “0001101”

50
 Digital Signaling
➢ Baud (Symbol Rate) :
D = N/T0 symbols/sec ; N- number of dimensions used in T0 sec.

➢ Bit Rate :
R = n/T0 bits/sec ; n- number of data bits sent in T0 sec.

Binary (2) Values Binary signal

wk
More than 2 Values Multilevel signal

➢ How to detect the data at the receiver?

1
w ( t )k* ( t ) dt ; k = 1,2...N
T0
wk =
Kk  0
Matched Filter Detection

w ( t ) − Waveform at the receiver input k* ( t ) − Orthogonal function

51
 Vector Representation
➢ Orthogonal function space corresponds to orthogonal vector
space :
N
w ( t ) =  wkk ( t ) 0  t  T0 Orthogonal Function Space
k =1
N

w =  w j j = ( w1 , w2 , w3 ,..., wN ) Orthogonal Vector Space

j =1

w is an N-dimensional vector, w = ( w1 , w2 , w3 ,..., wN )

  Orthogonal set of N-dimensional vectors

j

52
Vector Representation of a Binary Signal
➢ Examine the representation in next slide for the waveform of a 3-bit (binary)
signal. This signal can be directly represented by,
 
N =3
1   N =3    1 
s ( t ) =  d j p t −  j −  T  =  d j p j (t ) p j (t ) = p  t −  j −  T 
j =1   2   j =1   2 
The Pulse Shape p(t ),  p (t ) is not normalized
j

Vector d = d1 , d 2 , d3  = (1, 0,1) .

➢ Orthogonal function approach

M =3
s(t ) =  s j j ( t )  (t ) is the set of Orthonormal functions
j
j =1
 1
p j (t ) p j (t ) p j (t )  , ( j − 1)T  t  jT
 j (t ) = = = Or  j (t ) =  T j =1, 2, 3
p 2j ( t ) dt  0,
T0

Kj 25T
0  t Otherwise

The Orthonormal series coefficients are: s1 , s2 , s3 = (5 T , 0, 5 T )

53
 Vector Representation of a Binary Signal

A 3 bit Signal waveform

Vector Representation of the 3 bit signal

Bit shape pulse

Orthogonal Function Set

54
 Bandwidth Estimation
➢ The lower bound for the bandwidth of the waveform w(t) is given by the Dimensionality
Theorem
N 1  sinx  N 1
B = D (Hertz) Lower Bound for  k (t) →  , B = = D
2T0 2  x  2T0 2
➢ Binary Signaling:
N
Waveform: w(t ) =  wk k (t ) 0  t  T0 wk takes only BINARY values
k =1

Example: Binary signaling from a digital source: M=256 distinct messages

M = 2n = 28 = 256 ➔ Each message ~ 8-bit binary words
T0=8 ms – Time taken to transmit one message; Code word: 01001110
w1= 0, w2= 1, w3= 0, w4= 0, w5= 1, w6= 1, w7= 1, w8= 0

➢ Case 1: Rectangular Pulse Orthogonal Functions:

 k (t ) : unity-amplitude rectangular pulses;
T0
Tb = = 1 msec Time taken to send 1 bit of data
n
55
 Bandwidth Estimation (Binary Signaling)
➢ Receiver end: How are we going to detect data?
Orthogonal series coefficients wk are needed. Sample anywhere in the bit interval
1
w ( t )k* ( t ) dt ; k = 1,2...N
T0
wk =
Kk 
0
Matched Filter Detection

w ( t ) − Waveform at the receiver input k* ( t ) − Orthogonal function

n N
Bit Rate R = = 1 Kbits/s Baud Rate (Symbol Rate) D = = 1 Kbaud
T0 T0
Same as BINARY Signaling

1
The Lower Bound : B = D = 500 Hz
2
1
The actual Null Bandwidth: B = = D = 1000 Hz
Ts

Bandwidth: ➔ Null BW > lower bound BW

56
 Binary Signaling

Which wave shape gives lower bound BW? 57

 Binary Signaling
➢ Case 2: sin(x)/x Pulse Orthogonal Functions
Minimum Bandwidth
  
 sin  (t − kTs ) 
  Ts   Where T =T for the case of Binary signaling.
 k (t ) =    s b
 (t − kTs ) 
 Ts 

➢ Receiver end: How are we going to detect data?

Orthogonal series coefficients wk are needed. Sample at MIDPOINT of each interval

n N
Bit Rate R = = 1 Kbits/s Baud Rate (Symbol Rate) D = = 1 Kbaud
T0 T0
Same as BINARY Signaling
1
The ABSOLUTE Bandwidth: B = = 500 Hz LOWER BOUND bandwidth
2Ts

58
 Multilevel Signaling
N
Lower bound BW: B = For N=8 pulses, T0=8 ms => B=500Hz.
2T0

B Reduces, if N Reduces: So wk should take more than 2 values ( 2- binary signaling)

If wk’s have L>2 values → Resultant waveform – Multilevel signal

Multilevel data : Encoding l-bit binary data → into L-level : DAC

59
 Multilevel Signaling (Example)
M=256-message source ; L=4; T0=8 ms

Encoding Scheme: A 2-Bit Digital-to-Analog Converter

Binary Input Output Level
(l=2 bits) (V)
11 +3
10 +1
00 -1
01 -3
Binary code word - 01001110

w1= -3, w2= -1, w3= +3, w4= +1

n
Bit rate : R = =1 k bits/second
T0
Different
N 1
Baud ( symbol rate): D = = = 0.5 k baud
T0 Ts

Relation : R = lD Where l = log 2 ( L)

60
 Multilevel Signaling - Example

B=1/Ts=D=500 Hz

B=N/2T0=250Hz

➢ How can the data be detected at the receiver?

➢ Sampling at midpoint of Ts=2 ms interval for either case (T=1, 3, 5, 7 ms)
61
 Binary-to-multilevel polar NRZ Signal Conversion
➢ Binary to multilevel conversion is used to reduce the bandwidth required by the binary
signaling.
• Multiple bits (l number of bits) are converted into words having SYMBOL durations
Ts=lTb where the Symbol Rate or the BAUD Rate D=1/Ts=1/lTb.
• The symbols are converted to a L level (L=2l ) multilevel signal using a l-bit DAC.

• Note that now the Baud rate is reduced by l times the Bit rate R (D=R/l).
• Thus the bandwidth required is reduced by l times.
Ts: Symbol Duration L: Number of M ary levels
Tb: Bit Duration l: Bits per Symbol
L=2l D=1/Ts=1/lTb=R/l 62
Binary-to-multilevel Polar NRZ Signal Conversion

(c) L = 8 = 23 Level Polar NRZ Waveform Out

63
 Chapter 3:
Line Codes and Their Spectra
➢ Types of Line Codes
➢ Comparison of Line Codes
➢ PSD of Line Codes

64
 Line Codes in PCM
• The output of an ADC can be transmitted over a baseband channel.
• The digital information must first be converted into a physical signal.
• The physical signal is called a line code. Line coders use the terminology mark
to mean binary one and space to mean binary zero.
Analog
Input
Signal
Sample

X
Quantize ADC
XQ

Encode

Xk
Line
Code
x(t) PCM signal
65
 Line codes
1 1 0 1 0 0 1 BINARY DATA

(a) Punched Tape Mark Mark space Mark space space Mark
(hole) (hole) (hole) (hole)
Volts
A

0
(b) Unipolar NRZ
Tb Time

(c) Polar NRZ 0

-A

0
(d) Unipolar RZ

A
(e) Bipolar RZ 0
-A

A
(f) Manchester NRZ 0
-A

Binary Signaling Formats

66
 Goals of Line Coding
• A line code is designed to meet several goals:

– Self-synchronization.
• The ability to recover timing from the signal itself.
• Long series of ones and zeros could cause a problem.

– Low probability of bit error.

• The receiver needs to be able to distinguish the waveform associated
with a mark from the waveform associated with a space, even if there
is a considerable amount of noise and distortion in the channel.

– Spectrum that is suitable for the channel.

• In some cases DC components should be avoided if the channel has a
DC blocking capacitance.
• The transmission bandwidth should be minimized.

67
 Line Coder

Digital
Data a k
Line Coder
Physical x(t ) =
Waveform

k =−
ak p (t − kTb )

• The input to the line encoder is a sequence of values ak that is

a function of a data bit or an ADC output bit.
• The output of the line encoder is a waveform:

x(t ) = a
k =−
k p (t − kTb )

• Where p(t) is the Pulse Shape and Tb is the Bit Period

– Tb =Ts/n for n bit quantizer (and no parity bits).
– Rb =1/Tb=nfs for n bit quantizer (and no parity bits).
• The operational details of this function are set by the particular
type of line code that is being used.
68
 Types of Line Codes
• Each line code is described by a symbol mapping function ak
and a pulse shape p(t):

x(t ) = a
k =−
k p (t − kTb )

• Categories of line codes:

– Symbol mapping functions (ak).
• Unipolar
• Polar
• Bipolar (a.k.a. alternate mark inversion, pseudoternary)

– Pulse shapes p(t).

• NRZ (Nonreturn-to-zero)
• RZ (Return to Zero)
• Manchester (split phase)

69
 Unipolar NRZ Line Code
• The unipolar nonreturn-to-zero line code is defined by the
unipolar mapping:
+ A when X k = 1
ak = 
 0 when X k = 0
– where Xk is the kth data bit.
– In addition, the pulse shape for unipolar NRZ is:
 t 
p(t ) =    NRZ pulse shape
 Tb 
– Where Tb is the bit period.
Hard to recover symbol timing
when long string of 0s or 1s.
Note the DC component
This means wasted power!
1 0 1 1 0 1
A

0 Tb 2Tb 3Tb 4Tb 5Tb

70
 Unipolar RZ Line Code
• The unipolar return-to-zero line code has the same symbol
mapping but a different pulse shape than unipolar NRZ:

+ A when X k = 1
ak = 
 0 when X k = 0

 t 
p(t ) =    RZ pulse shape
 Tb / 2 
Long strings of 1s no longer a problem.
However strings of 0s still problem.
Pulse of half the duration of NRZ
requires twice the bandwidth!
1 0 1 1 0 1
A

0 Tb 2Tb 3Tb 4Tb 5Tb

71
 Polar Line Codes
• Polar line codes use the antipodal mapping:
+ A when X k = 1
ak = 
 − A when X k = 0
– Polar NRZ uses NRZ pulse shape.
– Polar RZ uses RZ pulse shape.
No DC component,
so more energy efficient.
1 0 1 1 0 1
A

Polar NRZ
Now we can handle
long strings of 0s, too.
A
Polar RZ

−A 72
 Manchester Line Codes
• Manchester line codes use the antipodal mapping and
the following split-phase pulse shape:
p(t )

 t + Tb / 4   t − Tb / 4 
p(t ) =    −   
 b T / 2   b T / 2 

1 0 1 1 0 1

• Easy synchronization and better spectral characteristics than polar RZ.

73
 Bipolar Line Codes
• With bipolar line codes a space is mapped to zero and a mark
is alternately mapped to -A and +A:
0 when X k = 0

ak = − A when X k = 1 and last mark → + A
+ A when X k = 1 and last mark → − A


• Also called pseudoternary signalling and alternate mark

inversion (AMI).
• Either RZ or NRZ pulse shape can be used.
1 0 1 1 0 1
A

Bipolar (RZ)

−A
74
 Comparison of Line Codes
• Self-synchronization:
– Manchester codes have built in timing information because
they always have a zero crossing in the center of the pulse.
– Polar RZ codes tend to be good because the signal level
always goes to zero for the second half of the pulse.
– NRZ signals are not good for self-synchronization.

• Error probability:
– Polar codes perform better (are more energy efficient) than
Unipolar or Bipolar codes.

• Channel characteristics:
– We need to find the PSD of the line codes to answer this ...

75
 Power Spectra for Binary Line Codes
➢ A digital signal is represented by

 t 
s(t ) = a n f (t = nTs ) ; f (t ) =    for unipolar NRZ
n =−  Ts 

f(t) - Symbol Pulse shape; Ts - Duration of one symbol;

Binary signaling : Ts= Tb , Multilevel signaling: Ts= lTb

➢ PSD depends on
(1) The pulse shape used
(2) Statistical properties of data expressed by the
autocorrelation function

76
 Power Spectra for Binary Line Codes
➢ PSD of a digital signal can be calculated using the
autocorrelation function:
2
−
F( f )
Ps ( f ) =
Ts
 R
k =
( k ) e − j 2 kfTs
Where { f (t )} = F ( f )
I
R(k ) =  (an an + k )i Pi The autocorrelation function of data
i =1

an and an + k are levels of the data pulses at the nth and (n+k)th symbol positions
Pi Probability of having the ith an an + k product

77
 PSD for Polar NRZ Signaling
Possible levels for the a’s : +A and -A
I
R(k ) =  (a a ) P a and a are the level of the pulses at the nth and ( n + k )th symbols
n n+k i i n n+k
i =1

2
1 1
R(0) =  (an an )i Pi = A2 + (− A) 2 = A2
i =1 2 2
4
For k  0, R(k ) =  (an an + k ) Pi = A21/ 4 + (− A)( A)1/ 4 + ( A)( − A)1/ 4 + ( − A) 21/ 4 = 0
i =1

 A2 , k = 0
 R polar (k ) = 
0, k  0
F( f )
2
sin  fTb 
2 kfTs
f (t ) =  (t / Tb )  F ( f ) = Tb
 fTb
Ps ( f ) =
Ts

k =−
R (k ) e

2
 sin  fTb 
PPolar NRZ ( f ) = A Tb 
2

  fTb 
78
 PSD for line codes
Unipolar NRZ

2
A2Tb  sin  fTb   1 
PUni. NRZ ( f ) =   1 +  ( f ) 
4   fTb   Tb 

Polar NRZ

2
 sin  fTb 
PPolar NRZ ( f ) = A2Tb  
  fTb 

Bit rate: R=1/Tb

79
 Chapter 3:
DIFFERENTIAL ENCODING

➢ Differential Encoding
➢ Eye Patterns
➢ Regenerative Receiver
➢ Bit Synchronizer
➢ Binary to Mary Conversion

80
 Differential Coding System
➢ Differential encoding removes the problem of Unintentional Signal Inversion.
➢ Polarity of the differentially encoded signal may be inverted without affecting the decoded
signal.

Modulo-2 addition
Exclusive OR

I1 I2 Out
0 0 0
0 1 1
1 0 1
1 1 0

en = d n  en −1
d n = en  en −1
 Represents Modulo-2 adder (XOR)

81
 Example of Differential Coding
Encoding en = d n  en−1
Input sequence dn 1 1 0 1 0 0 1
Encoded sequence en 1 0 1 1 0 0 0 1
Reference digit

~ ~ ~
Decoding (with correct channel polarity) d n = en  en−1
Receiver sequence en 1 0 1 1 0 0 0 1
(Correct polarity)
Decoded sequence dn 1 1 0 1 0 0 1

Decoding (with inverted channel polarity)

Received sequence en 0 1 0 0 1 1 1 0
(Inverted polarity)

Decoded sequence dn 1 1 0 1 0 0 1

➢ Decoded sequence is same whether there is inversion or not.

82
 Eye patterns
➢ The effects of channel filtering and channel noise can be seen by observing the received line
code on an oscilloscope.

Received Line
Code

Information from Eye Pattern

• Timing error → Eye opening
• Sensitivity → Slope of the
open eye
• Noise Margin → height of the
eye opening

83
 Regenerative Repeater
➢ Regenerate a noise-free digital signal. Amplify and clean-up the signal periodically

Produces a high level o/p

Increases the amplitude if sample value>VT
Produces a sample value

Minimize the effect of

channel noise & ISI
Generates a clocking signal
84
 Bit Synchronization
➢ Synchronization signals are clock-type signals necessary within a receiver for detection of
data from the corrupted input signal.
➢Digital communication need 3 types of synchronization signals.
• Bit Synchronization.
• Frame Synchronization.
• Carrier Synchronization.

Bit Synchronizer for NRZ Signals

85
Square-law Bit Synchronizer for NRZ Signals

86
 Binary-to-multilevel polar NRZ Signal Conversion
➢ Binary to multilevel conversion is used to reduce the bandwidth required by the binary
signaling.
• Multiple bits (l number of bits) are converted into words having SYMBOL durations
Ts=lTb where the Symbol Rate or the BAUD Rate D=1/Ts=1/lTb.
• The symbols are converted to a L level (L=2l ) multilevel signal using a l-bit DAC.

• Note that now the Baud rate is reduced by l times the Bit rate R (D=R/l).
• Thus the bandwidth required is reduced by l times.
Ts: Symbol Duration L: Number of M ary levels
Tb: Bit Duration l: Bits per Symbol
L=2l D=1/Ts=1/lTb=R/l Bnull=R/l 87
Power Spectra for Multilevel Polar NRZ Signals

(c) L = 8 = 23 Level Polar NRZ Waveform Out

88
 Spectral Efficiency
➢ The Spectral efficiency of a digital signal is given by, where R is the data rate and
B is the bandwidth required.
=
R ( Bit s )
B Hz
• If limited BW is desired, then use a signaling technique that has high spectral efficiency.
• Maximum spectral efficiency (which is limited by channel noise) is given by the
Shannon’s Channel Capacity formula:

C  S
max = = log 2 1 + 
B  N
Spectral efficiency for multilevel signaling is  =l
( bit s )
Hz

89
PSD of a multilevel polar NRZ waveform
I
R (k ) =  (an an + k )i Pi
i =1

For k = 0
8 1
R (0) =  (a n )i Pi = 21 where Pi =
2 for all of the eight possible values.
8
i =1

For k  0, R(k ) = 0.

Then the PSD for 2 (t ) is

F( f )
2

Pw2 ( f ) = (21 + 0) where the pulse width is Ts = 3Tb .

Ts
For the rectangular pulse width 3Tb : 2
F(f) −
Ps(f) =
Ts
 R ( k )e − j 2kfTs

PSD for a multilevel polar NRZ signal: k =

2
 sin lfTb 
Pmultilevel NRZ ( f ) = K  
 lfTb  where k is a constant
R
The null bandwidth is Bnull =
l
➢ Multilevel signaling is used to reduce the BW of a digital signal

90