Transcript William Stallings Data and Computer

William Stallings
Data and Computer
Communications
Chapter 5
Data Encoding
1
Data Communication Basics
Analog or Digital
Three Components
Data
Signal
Transmission
2
Analog Data Choices
analog line
analog voice
telephone
analog voice
digitized voice
digital line
Codec
01101000110
Codec: coder and decoder
3
Digital Data Choices
analog line
moduated data
modem
data
data
digital line
DSU
01101000110
DSU: data service unit
4
Encoding Techniques
Digital data, digital signal
Analog data, digital signal
Digital data, analog signal
Analog data, analog signal
5
Transmission Choices
Analog transmission
only transmits analog signals, without regard for data
content
attenuation overcome with amplifiers
Digital transmission
transmits analog or digital signals
uses repeaters rather than amplifiers
6
Advantages of Digital
Transmission
The signal is exact
Signals can be checked for errors
Noise/interference are easily filtered out
A variety of services can be offered over one
line
Higher bandwidth is possible with data
compression
7
Encoding schemes
Analog data, Analog signal
voice
analog
CODEC
analog
Modem
digital
analog
Telephone
Digital data, Analog signal
digital
Analog data, Digital signal
Digital data, Digital signal
digital
Digital
transmitter
digital
8
Encoding and Modulation
x(t)
x(t)
g(t)
digital
or
analog
Encoder
digital
g(t)
Decoder
t
s(f)
s(t)
m(t)
Modulator
digital
or
analog
m(t)
Demodulator
analog
fc
f
fc
9
Why encoding?
 Three factors determine successfulness of receiving a
signal
S/N (Signal to Noise Ratio)
data rate
bandwidth
10
Encoding Schemes' evaluation
factors
 Signal spectrum
 Clocking
 Error detection
 Signal interference & noise immunity
 Cost and complexity
11
Digital Data, Digital Signal /
Characteristics
Digital signal
Uses discrete, discontinuous, voltage pulses
Each pulse is a signal element
Binary data is encoded into signal elements
12
Terms (1)
Unipolar
All signal elements have same sign
Polar
One logic state represented by positive voltage the
other by negative voltage
Data rate
Rate of data transmission in bits per second
Duration or length of a bit
Time taken for transmitter to emit the bit
13
Terms (2)
Modulation rate
Rate at which the signal level changes
Measured in baud = signal elements per second
Mark and Space
Binary 1 and Binary 0 respectively
14
Interpreting Signals
Need to know
Timing of bits - when they start and end
Signal levels
Factors affecting successful interpretation of
signals:
Signal to noise ratio
Data rate
Bandwidth
15
Comparison of Encoding
Schemes (1)
Signal Spectrum
Lack of high frequencies reduces required bandwidth
Lack of dc component allows ac coupling via
transformer, providing isolation
It is important to concentrate power in the middle of
the bandwidth
Clocking issues
Synchronizing transmitter and receiver is essential
External clock is one way used for synchronization
Synchronizing mechanism based on signal is also
used & preferred (over using an external clock)
16
Mean square voltage per unit bandwidth
Spectral density
1.5
1
NRZ-L,
NRZI
B8ZS,HDB3
AMI, Pseudoternary
0.5
Manchester,
Differential Manchester
0
-0.5
0
0.5
1
1.5
Normalized frequency (f/r)
17
Comparison of Encoding
Schemes (2)
Error detection
Can be built into signal encoding
Signal interference and noise immunity
Some codes are better than others
Cost and complexity
Higher signal rate (& thus data rate) lead to higher
costs
Some codes require signal rate greater than data
rate
18
Encoding Schemes
Nonreturn to Zero-Level (NRZ-L)
Nonreturn to Zero Inverted (NRZI)
Bipolar -AMI (Alternate Mark Inversion)
Pseudoternary
Manchester
Differential Manchester
B8ZS
HDB3
19
Digital data, Digital signal
0 1 0 0
1 1 0 0
0 1 1
NRZ
NRZI
Bipolar -AMI
Pseudoternary
Manchester
Differential
Manchester
20
Nonreturn to Zero-Level (NRZ-L)
Two different voltages for 0 and 1 bits
Voltage constant during bit interval
Most often, negative voltage for one value and
positive for the other
21
Nonreturn to Zero Inverted
Nonreturn to zero inverted on ones
Constant voltage pulse for duration of bit
Data encoded as presence or absence of signal
transition at beginning of bit time
Transition (low to high or high to low) denotes a
binary 1
No transition denotes binary 0
An example of differential encoding (Data
represented by changes rather than levels)
22
NRZ
23
NRZ pros and cons
Pros
Easy to engineer
Makes good use of bandwidth
Cons
dc component
Lack of synchronization capability
Used for magnetic recording
Not often used for signal transmission
24
Multilevel Binary
Use more than two levels
Bipolar-AMI
zero represented by no line signal
one represented by positive or negative pulse
one pulses alternate in polarity
No loss of sync if a long string of ones happens
(zeros still a problem)
No net dc component  Can use a transformer for
isolating transmission line
Lower bandwidth
Easy error detection
25
Pseudoternary
One represented by absence of line signal
Zero represented by alternating positive and
negative
No advantage or disadvantage over bipolar-AMI
26
Bipolar-AMI and Pseudoternary
27
Trade Off for Multilevel Binary
Not as efficient as NRZ
With multi-level binary coding, the line signal may
take on one of 3 levels, but each signal element,
which could represent log23 = 1.58 bits of
information, bears only one bit of information
Receiver must distinguish between three levels
(+A, -A, 0)
Requires approx. 3dB more signal power for same
probability of bit error
28
Biphase
Manchester
Transition in middle of each bit period
Transition serves as clock and data
Low to high represents one
High to low represents zero
Used by IEEE 802.3 (Ethernet)
Differential Manchester
Midbit transition is for clocking only
Transition at start of a bit period represents zero
No transition at start of a bit period represents one
Note: this is a differential encoding scheme
Used by IEEE 802.5 (Token Ring)
29
Biphase Pros and Cons
Con
At least one transition per bit time and possibly two
Maximum modulation rate is twice NRZ
Requires more bandwidth
Pros
Synchronization on mid bit transition (self clocking)
No dc component
Error detection
Absence of expected transition points to error in
transmission
30
Modulation Rate
R=Data Rate=bits/sec=1
Mbps for both cases
Modulation Rate=Baud
Rate=Rate at which signal
elements are generated=R for
NRZI=2R for Manchester
31
Scrambling Techniques
 Used to reduce signaling rate relative to the data
rate by replacing sequences that would produce
constant voltage for a priod of time with a filling
sequence that accomplishes the following goals:
Must produce enough transitions to maintain syncchronization
Must be recognized by receiver and replaced with original data
sequence
is same length as original sequence
 No dc component
 No long sequences of zero level line signal
 No reduction in data rate
 Error detection capability
 As an example, fax machines use the modified
Huffman code to accomplish this.
32
B8ZS
Bipolar With 8 Zeros Substitution
Based on bipolar-AMI
If octet of all zeros and last voltage pulse
preceding was positive, encode as 000+-0-+
If octet of all zeros and last voltage pulse
preceding was negative, encode as 000-+0+Causes two violations of AMI code
This is unlikely to occur as a result of noise
Receiver detects and interprets the sequence as
octet of all zeros
33
HDB3
High Density Bipolar 3 Zeros
Based on bipolar-AMI
String of four zeros replaced with one or two
pulses
Note: The following is the explanation for the HDB3 code example on the
next slide (see rules in Table 5.4, page 142):
Assuming that an odd number of 1's have occurred since the last substitution,
since the polarity of the preceding pulse is "-", then the first 4 zeros are
replaced by "000-". For the next 4 zeros, since there have been no Bipolar
pulses since the 1st substitution, then they are replaced by"+00+" since the
preceding pulse is a "-". For the 3rd case where 4 zeros happen, 2 (even)
Bipolar pulses have happened since the last substitution and the polarity of
the preceding pulse is "+", so "-00-" is substituted for the zeros.
34
B8ZS and HDB3
(Assume odd number of 1s
since last substitution)
See Table 5.4 for HDB3 Substitution Rules
35
Digital Data, Analog Signal
Transmitting digital data through PSTN (Public
telephone system)
300Hz to 3400Hz bandwidth
modem (modulator-demodulator) is used to convert
digital data to analog signal and vice versa
Three basic modulation techniques are used:
Amplitude shift keying (ASK)
Frequency shift keying (FSK)
Phase shift keying (PSK)
36
Modulation Techniques
37
Amplitude Shift Keying
Values represented by different amplitudes of
carrier
Usually, one amplitude is zero
i.e. presence and absence of carrier is used
Susceptible to sudden gain changes
Inefficient
Up to 1200bps on voice grade lines
Used over optical fiber
38
ASK
Vd(t)
Vc(t)
VASK(t)
Signal
power
frequency spectrum
fc-3f0 fc-f0
fc fc+f0
fc+3f0
Frequency
39
Frequency Shift Keying
Values represented by different frequencies
(near carrier)
Less susceptible to error than ASK
Up to 1200bps on voice grade lines
High frequency radio (3-30 MHz)
Higher frequency on LANs using co-ax
40
FSK
Data
signa
l
Carrier 1
v1(t)
Carrier 2
v2(t)
vd(t)
vFSK(t)
Signal
power
frequency spectrum
Frequency
f1
f2
41
FSK in modem (on Voice Grade
Line)
Amplitude
PSTN bandwidth
400
1180 1650
(1270) (2025)
1850
980
(2225)
(1070)
3400
Frequency(Hz)
frequency spectrum
42
Phase Shift Keying
Phase of carrier signal is shifted to represent
data
Differential PSK
Phase shifted relative to previous transmission rather
than some reference signal
43
PSK
Data
Signal
vc(t)
Carrier vc(t)
Phase
coherent vPSK(t)
Differential
v’PSK(t)
bit rate = signaling rate
Differential example: for every logic 1,
180 degree phase shift
180=0
0=1
phase diagram
44
Quadrature PSK
More efficient use by each signal element
representing more than one bit
e.g. shifts of /2 (90o)
Each element represents two bits
Can use 8 phase angles and have more than one
amplitude
9600bps modems use 12 angles , four of which have
two amplitudes
45
Multilevel modulation method
00
0°
01
10
11
+90° +180°
+270°
bit rate = n x signaling rate
46
Multilevel modulation method
+90°=01
+180°=10
0°=00
+270°=11
16-QAM phase diagram
4-PSK phase diagram
47
Performance of Digital to
Analog Modulation Schemes
Bandwidth
ASK and PSK bandwidth directly related to bit rate
FSK bandwidth related to data rate for lower
frequencies
requires more analog bandwidth than ASK
(See Stallings for math)
In the presence of noise, bit error rate of PSK
and QPSK are about 3dB superior to ASK and
FSK
48
Analog Data, Digital Signal
Digitization
Conversion of analog data into digital data
Digital data can then be transmitted using NRZ-L or
using other codes
Digital data can then be converted to analog signal
Analog to digital conversion done using a codec
Pulse code modulation
Delta modulation
49
Analog data, Digital signal
 Two principle techniques used
PCM (Pulse Code Modulation)
DM (Delta Modulation)
Sampling
clock
Analog
voice
signal
Sampling
Circuit
PAM signal
PCM signal
Quantizer
and compander
Digitized
voice
signal
50
Pulse Code Modulation(PCM) (1)
If a signal is sampled at regular intervals at a
rate higher than twice the highest signal
frequency, the samples contain all the
information of the original signal
(Proof - Stallings appendix 4A)
Voice data limited to below 4000Hz
Require 8000 sample per second
Analog samples (Pulse Amplitude Modulation,
PAM)
Each sample assigned digital value
51
Pulse Code Modulation(PCM) (2)
4 bit system gives 16 levels
Quantized
Quantizing error or noise
Approximations mean it is impossible to recover
original exactly
8 bit sample gives 256 levels
Quality comparable with analog transmission
8000 samples per second of 8 bits each gives
64kbps
52
The process starts with an analog signal, which is
sampled by PAM sample. the resulting pulse are
quantized to produced PCM pulses and then encoded
to produce bit stream. At the receiver end, the process
is reversed to reproduce the analog signal.
53
PCM
 Sampling signal based on nyquist theorem
Original signal
PAM pulse
PCM pulse
with quantized error
3.2
3
4
2.8
3.4
3
3
4.2
1.2
4
1
011
PCM output
3.9
100
011
011
001
100
011100011011001100
54
Nonlinear Encoding
Quantization levels are not necessarily equally
spaced. The problem with equal spacing is that the
mean absolute error for each sample is the same,
regardless the signal level. Lower amplitude values
are relatively more distorted.
Nonlinear encoding reduces overall signal
distortion
Can also be done by companding
55
Nonlinear encoding
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Quantizing level
15
14
13
Strong signal
Weak signal
12
11
10
9
8
76
5
4
3
2
1
0
Without nonlinear encoding With nonlinear encoding
56
Prior to the input signal being sampled and converted by ADC into a
digital form, it is passed through a circuit known as a compressor.
Similarly, at the destination, the reverse operation is perform on the
output of the DAC by a circuit known as expander.
57
Delta Modulation
Analog input is approximated by a staircase
function
Move up or down one level () at each sample
interval
Binary behavior
Function moves up or down at each sample interval
58
Delta Modulation - example
59
Delta Modulation - Performance
Good voice reproduction
PCM - 128 levels (7 bit)
Voice bandwidth 4khz
Should be 8000 x 7 = 56kbps for PCM
Data compression can improve on this
e.g. Interframe coding techniques for video
60
Analog Data, Analog Signals
Why modulate analog signals?
Higher frequency can give more efficient
transmission
Permits frequency division multiplexing (chapter 8)
Types of modulation
Amplitude
Frequency
Phase
61
Analog
Modulation
62
Spread Spectrum
Analog or digital data
Analog signal
Spread data over wide bandwidth
Makes jamming and interception harder
2 schemes are used:
Frequency hoping
Signal broadcast over seemingly random series of
frequencies
Direct Sequence
Each bit is represented by multiple bits in transmitted
signal known as a chipping code
63