Beruflich Dokumente
Kultur Dokumente
Telecomunicazioni
Docente: Andrea Baiocchi
Dip. INFOCOM - Stanza 35, 1 piano palazzina P. Piga Sede Facolt S. Pietro in Vincoli E-mail: andrea.baiocchi@uniroma1.it
Programma
1. 2. 3. 4. 5. 6. 7.
SERVIZI E RETI DI TELECOMUNICAZIONE FONDAMENTI DI COMUNICAZIONI ARCHITETTURE DI COMUNICAZIONE MODI DI TRASFERIMENTO LO STRATO DA ESTREMO A ESTREMO: UDP E TCP LO STRATO DI RETE IN INTERNET TECNOLOGIE DI STRATO DI COLLEGAMENTO
Digital Representation of Information Digital Representation of Analog Signals Chapter 3 Why Digital Communications? Characterization of Communication Channels Fundamental Limits in Digital Transmission Line Coding Modems and Digital Modulation Properties of Media and Digital Transmission Systems Error Detection and Correction
Digital Networks
!
Telephone
Questions of Interest
! ! ! !
Can we reduce all information to sequences of bits? How? How many bits do we need to represent a message (text, speech, image)? How fast does the network/system transfer information? Under which quality constraints? How can we deal with errors?
! !
How are errors introduced? How are errors detected and corrected?
What transmission speed and coding of data is possible over radio, copper cables, fiber, infrared, ?
Fundamentals of communications
Digital Representation of Information
Adapted from slides of the book: A. Leon Garcia, I. Widjaja, Communication networks, McGraw Hill, 2004
Bit: BInary digiT Either symbol belonging to a set of two elements or number with value 0 or 1
! ! !
n bits: digital representation for 0, 1, , 2n1 Byte or Octet, n = 8 Computer word, typically n = 32, or 64 n-bit field in a header n-bit representation of a voice sample Message consisting of n bits
The number of bits required to represent a message is a measure of its information content
!
Stream ! Information that is produced and possibly conveyed over a communication system continuously
! !
Format ASCII A4 page 200x100 pixels/in2 8x10 in2 photo 4002 pixels/in2
Color Image
W H Color image = H W
Red component image
W + H
Green component image
W + H
Blue component image
Stream Information
! !
A real-time voice signal must be digitized and transmitted or recorded as it is produced Analog signal level varies continuously in time
Th e s p ee
ch s
g n al l e
v el
v a r ie s w i th
m(e)
Analog signal
!
In communications engineering signal refers to e physically measurable entity that can be used to carry information
!
E.g. e.m. field, voltage and current in lumped circuits, air pressure
An analog signal is a function x(t) defined over the real axis and taking values in an interval of the real line Information is carried by the values of x(t) at each time t
Images are natively carried by analog signals, under the form of e.m. field in the visible bandwidth, i.e. the range of frequencies that produces a reaction in human sight sensors Sounds are natively carried by analog signals, i.e. variation of air pressure that produces a reaction in human hearing sensors Analog nature is due to the fact that for our purposes all these phenomena are well described by classical physics models (Newton mechanics, Maxwell e.m. field theory) and classical physics rests on classical analysis to describe its models (variables taking values on a continuum, e.g. real axis).
3 bits / sample
(a) Original 7"/2 waveform and 5"/2 the sample 3"/2 values "/2 -"/2 -3"/2 -5"/2 -7"/2 (b) Original waveform and the quantized values 7"/2 5"/2 3"/2 "/2 -"/2 -3"/2 -5"/2 -7"/2
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
Figure 3.2
CD Audio
! ! ! !
Cellular phones use more powerful compression algorithms: e.g. 6.5-13 kbps for GSM
Ws = 22 kHertz -> 44000 samples/sec 16 bits/sample Rs=16 x 44000= 704 kbps per audio channel MP3 uses more powerful compression algorithms: 50 kbps per audio channel
Video Signal
!
Each picture digitized & compressed 10-30-60 frames/second depending on quality Small frames for videoconferencing Standard frames for conventional broadcast TV HDTV frames
Frame resolution
!
30 fps
Video Frames
176 QCIF videoconferencing 144 720 Broadcast TV 480 1920 HDTV 1080
at 30 frames/sec = 67 x 106 pixels/sec at 30 frames/sec = 10.4 x 106 pixels/sec at 30 frames/sec = 760,000 pixels/sec
176x144 or 352x288 pix @10-30 fr/sec Full Motion MPEG2 720x480 pix @30 fr/sec HDTV MPEG2 1920x1080 @30 fr/sec
Constant bit-rate
! !
Variable bit-rate
!
Signals such as digitized telephone voice produce a steady stream: e.g. 64 kbps Network must support steady transfer of information, e.g. 64 kbps circuit Signals such as digitized video produce a stream that varies in bit rate, e.g. according to motion and detail in a scene Network must support variable transfer rate of information with possibly a guaranteed minimum rate, e.g. packet switching with traffic engineering functions
! !
Loss: Is information delivered without loss? If loss occurs, is delivered signal quality acceptable?
! ! !
Applications & application layer protocols developed to deal with these impairments
Transmission Delay
L R L/R tprop d c
number of bits in message bps speed of digital communication system time to transmit the message time for signal to propagate across medium distance in meters speed of light (3x108 m/s in vacuum)
seconds
Use data compression to reduce L Use higher speed modem to increase R Place far end system closer to reduce d
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
Compression
! !
Represent the information using fewer bits than provided natively Noiseless: original information recovered exactly
!
Compression Ratio
Fundamentals of communications
Digital Representation of Analog Signals
Adapted from slides of the book: A. Leon Garcia, I. Widjaja, Communication networks, McGraw Hill, 2004
"
"
! !
Sampling: obtain samples of x(t) at uniformly spaced time intervals: xk=x(tk), tk=t0+kT, k integer T is the sampling time, F=1/T is the sampling rate. Quantization: map each sample xk into an approximation value yk=f(xk) of finite precision Compression: to lower bit rate further, apply additional compression method
Differential coding: cellular telephone speech Subband coding: MP3 audio
A signal that varies faster needs to be sampled more frequently Bandwidth measures how fast a signal varies
1 01 01 01 0 11 1 1 0 000 ...
t
...
...
...
t
1 ms
! !
1 ms
Periodic Signals
!
A periodic signal with period T can be represented as sum of sinusoids using Fourier Series: x(t) = a0 + a1cos(2!f0t + "1) + a2cos(2$2f0t + "2) + + akcos(2!kf0t + "k) +
DC long-term average
kth harmonic
|ak|2 determines amount of power in kth harmonic Amplitude specturm |a0|, |a1|, |a2|,
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
x2(t) ...
11 1 1 0 000 ...
t
T2 =0.25 ms
T1 = 1 ms
Spectrum of x1(t) Spectrum of a signal: magnitude of amplitudes as a function of frequency x1(t) varies faster in time & has more high frequency content than x2(t) Bandwidth W is defined as Spectrum of x2(t) range of frequencies where a signal has non-negligible power, e.g. range of band that contains 99% of total signal power
1.2 1 0.8 0.6 0.4 0.2 0
18
21
24
27
15
12
frequency (kHz)
1.2 1
30
33
36
39 39
18
30
15
12
frequency (kHz)
24
27
33
36
42
21
42
(noisy )
|p
(air stopped)
| ee (periodic)
| t (stopped) | sh (noisy)
! !
Not all signals are periodic ! E.g. voice signals varies according to sound X(f) ! Vowels are periodic, s is noiselike Spectrum of long-term signal ! Averages over many sounds, many speakers ! Involves Fourier transform Telephone speech: 4 kHz CD Audio: 22 kHz
f 0 W
Sampling theorem
Sampling theorem (Nyquist): Perfect reconstruction if sampling rate 1/T ! 2W
(a)
x(t) t x(nT) t
Sampler
(b)
x(nT) t Interpolation filter x(t) t
input x(nT)
Uniform quantizer
Quantizer maps input into closest of 2m representation values Quantization error: noise = y(nT) x(nT)
Original signal Sample value Approximation
Quantizer error
M = 2m quantization levels Dynamic range (V, V) Quantization interval " = 2V/M (uniform quantization)
y(nT) ... #2" "
!
3 bits / sample
! 2
-V
" 2
V x(nT)
If the number of levels M is large, the error e(x)=y(x)x is approximately uniformly distributed between ("/2, "/2) in each quantization interval
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
Quantizer performance
!
The SNR is usually stated in decibels: SNR dB = 10 log10(#x2/#e2) = 6m + 10 log10(3#x2/V2) Example: SNR dB = 6m 7.27 dB for V/#x = 4.
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
2W samples / sec
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
Assume V/#x = 4, then 40 dB = 6m 7.27 & m = 8 bits/sample PCM (Pulse Code Modulation): Bit rate= 8000 x 8 bits/sec= 64 kbps
Fundamentals of communications
Why Digital Communications?
Adapted from slides of the book: A. Leon Garcia, I. Widjaja, Communication networks, McGraw Hill, 2004
A Transmission System
Transmitter Communication channel Transmitter
!
!
Receiver
Receiver
! !
Signal = measurable physical quantity that can be modified according to the value of the data to be transmitted, conveyed over a transmissin medium and detected by a receiving device.
Receives energy from medium Converts received signal into form suitable for delivery to user
Transmission Impairments
Transmitter
Transmitted Signal Received Signal Receiver
Communication channel
! ! ! ! !
Including infrared
Digital transmission: only discrete levels need to be reproduced Sent Distortion Attenuation Received
...
Repeater
Destination
Attenuation is removed (amplifier) Distortion is not completely eliminated In-band noise & interference can be removed only in part (out of band)
Signal quality decreases with # of repeaters Attenuated and distorted signal + noise Recovered signal + residual noise
Repeater
Amp Equalizer
...
Regenerator
Destination
Regenerator recovers original data (bit) sequence from degraded signal and retransmits on next segment by using a clean signal
!
2T
-A
3T
4T
5T
6T
Example of clock
Clock signal
0 1 0 1 1 0 1 0 0 0 1 1 0
1 0
Message bits
+d -d
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
Up 8 Mbps down (ADSL) Coexists with analog telephone 20 Mbps down (ADSL2+) signal 50 Mbps down (VDSL) From 1 to 54 Mbps IEEE 802.11b/a/g wireless LAN 1.5-45 Mbps 2.5-10 Gbps >1600 Gbps 5 km multipoint radio 1 wavelength Many wavelengths
Fundamentals of communications
Characterization of Communication Channels
Adapted from slides of the book: A. Leon Garcia, I. Widjaja, Communication networks, McGraw Hill, 2004
Communications Channels
!
Communications system includes electronic or optical devices that are part of the path followed by a signal
!
Transmitter, equalizers, amplifiers, filters, couplers, detector, clocks and carrier generators
By communication channel we refer to the combined end-to-end physical medium and attached devices
Information source Channel Information destination Information rendering Final user Demodulation, equalization, symbol decision
Channel decoder
Line decoder
Speed: Bit rate, R bps Reliability: Bit error rate, BER=10k Focus of this section
Communications Channel
Transmitted Signal Received Signal
Transmitter
Communication channel
Receiver
Bandwidth ! In order to transfer data faster, a signal has to vary more quickly. ! A channel or medium has an inherent limit on how fast the signals it passes can vary ! Channel bandwidth limits how tightly input pulses can be packed
Impairments ! Signal attenuation ! Signal distortion ! Spurious noise ! Interference from other signals ! Channel impairments limit accuracy of measurements on received signal
We often assume two basic properties of channels Linearity y1(t)=Ch[x1(t)], y2(t)=Ch[x2(t)] => y1(t)+y2(t)=Ch[x1(t)+x2(t)]
!
t A(f) =
!
Aout Ain
Linear: superposition of effects holds Stationary: input-output relationship does not vary over time Output is sinusoid at same frequency, but attenuated & phaseshifted Sinusoids are autofunctions of LTI systems
Measure amplitude of output sinusoid (of same frequency f) and calculate amplitude response A(f) = ratio of output amplitude to input amplitude
! !
If A(f) ! 1, then input signal passes readily If A(f) ! 0, then input signal is blocked
Ideal filter: all sinusoids with frequency f<Wc are passed without attenuation and delayed by ( seconds; sinusoids at other frequencies are blocked y(t) = Aincos(2$ft 2$f() = Aincos(2$f(t ( )) = x(t()
Amplitude Response 1 0 Phase Response
'(f) = 2$f(
1/ 2$ f
Wc
Amplitude Response 1
Phase Response
A(f) = (1+4$2b2f2)1/2
0 -45o
$(f) = arctan(2$bf)
1/ 2$ f
1/" 2
Wc
-90o
f1
!
f2 f
Some channels pass signals within a band that excludes low frequencies
!
! !
Channel bandwidth is the width of the frequency band that passes non-negligible signal power Example. 3dB bandwidth: frequency interval where output power density is no less than 1/2 than peak value
Channel Distortion
x(t) = )k akcos(2$fkt + *k)
! !
Channel
y(t)
Let x(t) be a digital signal bearing data information How well does y(t) follow x(t)?
If amplitude response is not flat, then different frequency components of x(t) will be transferred by different amounts If phase response is not linear, then different frequency components of x(t) will be delayed by different amounts
0 0
0 0
1
...
f0=1/T=1000 Hz t
Let x(t) input to ideal lowpass filter that has zero delay and Wc = 1.5 kHz, 2.5 kHz, or 4.5 kHz
! ! !
Wc = 1.5 kHz passes only the first two terms Wc = 2.5 kHz passes the first three terms Wc = 4.5 kHz passes the first five terms
Amplitude Distortion
0.125 0.25 0.625 0.75 0.375 0.875 0.5 0 1
0.125
0.375
0.625
0.75
0.875
0.25
As the channel bandwidth increases, the output of the channel resembles the input more closely
0.5
0.375
0.625
0.75
0.125
0.875
0.25
0.5
Time-domain Characterization
+(t) a Channel h(t) t
0
! ! !
td
Time-domain characterization of a channel requires finding the impulse response h(t) Apply a very narrow pulse of amplitude a to a channel at time ( and observe the channel output at time t The output in case of a linear, stationary, causal channel is y(t) = 0, t < ( y(t) = ah(t(), t > (
Impulse response
!
a+(tTa)+b+(tTb) ah(tTa)+bh(tTb) If a signal x(t) is applied to LTI channel input, then y(t) = " x(()h(t()d(
!
x(t) can be thought of as the sum of pulses at times ( (#< (<#) with amplitude x((), so y(t) is the sum of responses x(()h(t()
It can be shown that H(f) is the Fourier transform of h(t) and thatY(f) = H(f)X(f)
Output
xk+(tkT)
#k xk+(tkT)
xkh(tkT)
#k xkh(tkT)
In case of pulse transmission in a linear, time invariant, additive noise channel, we get an output signal y(t): noise
yn = y(nT) = $k xk hnk + zn
with hnk = h(nTkT) and zn = z(nT).
output sample
!
ISI is the undesired interference coming from tails of pulses other than the n-th one and giving non-null contributions to the n-th output sample
Given a channel response H(f), we can add reception and possibly transmission filters, so that the overall (filtered) channel response is Hc(f) = HTX(f) H(f) HRX(f) To get null ISI at sampling rate 1/T it must be
$k Hc(fk/T) = cost
!
For a low-pass channel with Hc(f)=0 for |f|>Wc this is not possible unless 1/T <= 2Wc.
For channel with ideal low-pass amplitude response of bandwidth Wc, the impulse response is a Nyquist pulse h(t)=s(t(), where T =1/(2Wc), and
1.2 1 0.8 0.6 0.4 0.2 0 -1 T 0 -0.2 -0.4
-7T
-6T
-5T
-4T
-3 T
-2 T
1T
2T
3T
4T
5T
6T
7T
! !
s(t) has zero crossings at t = kT, k = 1, 2, Pulses can be packed every T seconds with zero InterSymbol Interference (ISI)
0 -2 T -1T 0 1T 2T 3T
4T
-1
-s(t-2T)
r(t)
2 1 0
-2T
-1T -1 -2
1T
2T
3T
4T
If channel is ideal low pass with Wc, then pulses maximum rate pulses can be transmitted without ISI is 2Wc pulse/s s(t) is one example of class of Nyquist pulses with zero ISI
!
Problem: sidelobes in s(t) decay as 1/t which add up quickly when there are slight errors in timing Requires slightly more bandwidth than Wc Sidelobes decay as 1/t3, so more robust to timing errors
A(f)
(1,)Wc Wc
(1+,)Wc
Eye diagram
!
10 Gbit/s signal after transmission through a dispersive channel (with non negligible ISI)
Fundamentals of communications
Fundamental Limits in Digital Transmission
Adapted from slides of the book: A. Leon Garcia, I. Widjaja, Communication networks, McGraw Hill, 2004
! !
p(t) pulse at receiver in response to a single input pulse (takes into account pulse shape at input, transmitter & receiver filters, and communications medium) r(t) waveform that appears in response to sequence of pulses If s(t) is a Nyquist pulse, then r(t) has zero intersymbol interference (ISI) when sampled at multiples ofT
1 +A -A
0
0
T
1
2T
1
3T
0
4T
1
5T t
r(t) Receiver
Transmitter Filter
Communication Medium
Receiver Filter
Received signal
Multilevel Signaling
!
Nyquist pulses achieve the maximum signalling rate with zero ISI
!
With two signal levels, each pulse carries one bit of the source bit stream
!
In the absence of noise, the bit rate can be increased without limit by increasing m BUT Additive noise limits # of levels that can be used reliably.
Four levels {-1, -1/3, 1/3, +1} for {00,01,10,11} Waveform for 11,10,01 sends +1, +1/3, -1/3 Zero ISI at sampling instants
1.2 1 0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -1 0 1 2 3
Composite waveform
! ! !
Presence of noise limits accuracy of measurement of received signal amplitude Errors occur if signal separation is comparable to noise level Noise places a limit on how many amplitude levels can be used in pulse transmission
Receiver makes decision based on (sampled) received signal level = source pulse level + noise
!
Error rate depends on relative value of noise amplitude and spacing between signal levels Large (positive or negative) noise values can cause wrong decision
+A
+A +5A/7 +3A/7 +A/7 -A/7 Typical noise -3A/7 -5A/7 Eight signal levels -A
+A/3
-A/3
-A
Noise
!
! !
Thermal electronic noise is inevitable (due to vibrations of electrons); thermal Noise can be modeled as a white Gaussian process
! !
Often interference from a large number of scattered and similar sources can be modeled as white Gaussian noise
Gaussian noise
x
x0 %2 = Avg Noise Power t
1 2" !
e#x
2 2! 2
x0
Probability of Error
!
Error occurs if noise value exceeds the information signal magnitude over the decision threshold With two-level signalling, +A and A, probability of error is Q(A/%)
0
1.00E+00 1.00E-01 1.00E-02 1.00E-03 1.00E-04 1.00E-05 1.00E-06 1.00E-07 1.00E-08 1.00E-09 1.00E-10 1.00E-11 1.00E-12
Q(x)
Role of SNR
! ! !
With M=2m levels per symbol, the tx symbol values are ak=A+(2k1)A/M, with k=1,,M. With equiprobable symbols: E[ak2]=(M21)A2/(3M2)=PTX Received sample is (dispersive channel, zero ISI): yn=h0xn+zn.
noise
error-prone SNR = Average Signal Power Average Noise Power SNR (dB) = 10 log10 SNR
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
Cb = Wc log2(1+SNR)
bit/s
Bandlimited channel (Wc) X Y=X+Z 2Wc symbols/s Input symbol Output symbol (Gaussian) Z White gaussian noise
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
It can be shown that in case of real input signal the optimal source is gaussian and the AWGN capacity is C = 0.5 log2(1+P/PN) [bit/symbol] where PN is the additive noise power, P is the useful signal received power A dispersive, additive noise channel can be reduced to AWGN if zero ISI is provided; to that end it must be symbol rate < 2Wc. Then Cb = Wclog2(1+(Eb/N0)Rb/Wc) [bit/s]
Reliable communications is possible if the tx rate Rb<Cb. If Rb > Cb, then reliable communications is not possible.
Reliable means the bit error rate (BER) can be made arbitrarily small through sufficiently complex coding.
!
Cb can be used as a measure of how close a system design is to the best achievable performance. P= average power of input signal N0=noise power spectral density=k-F, k=1.381023 J/K, -=noise temperature, typically 300 K, F=noise figure, typically 6 dB
SNR=P/(N0Wc), with
! !
Example
!
Find the Shannon channel capacity for a telephone channel with Wc = 3400 Hz and SNR = 10000
Fundamentals of communications
Line Coding
Adapted from slides of the book: A. Leon Garcia, I. Widjaja, Communication networks, McGraw Hill, 2004
Mapping of binary information sequence into the digital signal that enters the channel
!
Transmitted power: Power consumption = $ Bit timing: Transitions in signal help timing recovery Bandwidth efficiency: Excessive transitions wastes bw Low frequency content: Some channels block low frequencies
! !
long periods of +A or of A causes signal to droop Waveform should not have low-frequency content
! !
Error detection: Ability to detect errors helps Complexity/cost: Is code implementable in chip at high speed?
NRZ-inverted (differential encoding) Bipolar encoding Manchester encoding Differential Manchester encoding
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
Figure 3.35
NRZ Bipolar
!
power density
NRZ has high content at low frequencies Bipolar tightly packed around T/2 Manchester wasteful of bandwidth
Manchester
0.2
0.4
0.6
0.8
1.2
1.4
1.6
-0.2
1.8
fT
Polar NRZ ! 1 maps to +A/2 pulse ! 0 maps to A/2 pulse ! Better Average Power
! ! ! !
Bipolar Code
1 Bipolar Encoding
! ! !
Every +pulse matched by pulse so little content at low frequencies Spectrum centered at T/2
! !
1 maps into A/2 first T/2, A/2 last T/2 0 maps into -A/2 first T/2, A/2 last T/2 Every interval has transition in middle
! !
! ! ! ! ! !
! !
mBnB line code Maps block of m bits into n bits Manchester code is 1B2B code 4B5B code used in FDDI LAN 8B10B code used in Gigabit Ethernet 64B66B code used in 10G Ethernet
Differential Coding
1
NRZ-inverted (differential encoding) Differential Manchester encoding
!
Errors in some systems cause transposition in polarity, +A become A and vice versa ! All subsequent bits in Polar NRZ coding would be in error Differential line coding provides robustness to this type of error
! !
! ! !
Same spectrum as NRZ Errors occur in pairs Also used with Manchester coding
1 mapped into transition in signal level 0 mapped into no transition in signal level
Fundamentals of communications
Modems and Digital Modulation
Adapted from slides of the book: A. Leon Garcia, I. Widjaja, Communication networks, McGraw Hill, 2004
Bandpass Channels
fc Wc/2 fc fc + Wc/2
0
!
Digital modulators embed information into waveform with frequencies passed by bandpass channel
!
A sinusoidal signal with frequency fc centered in middle of bandpass channel is named carrier Modulators embed information into the carrier
-1
Map bits into amplitude of sinusoid: 1 send sinusoid; 0 no sinusoid Demodulator looks for signal vs. no signal
-1
2T
3T
4T
5T
Map bits into frequency: 1 send frequency fc + + ; 0 send frequency fc - + Demodulator looks for power around fc + + or fc - +
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
Phase Modulation
Information
Phase Shift Keying +1
T 2T 3T 4T 5T 6T
-1
!
Modulator
Modulate cos(2$fct) by multiplying by Ak for T seconds:
Ak
Demodulator
Demodulate (recover Ak) by multiplying by 2cos(2$fct) for T seconds and lowpass filtering (smoothing):
Y(t) = Akcos(2$fct) Received signal during kth interval
x
2cos(2$fct)
X(t)
Multiplication for the local carrier and integration over a symbol interval takes place for every symbol Synchronous demodulation (perfect local carrier)
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
Example of Modulation
Information Baseband Signal
1 +A -A +A -A
0 T 0 T
2T
3T
4T
5T
6T
2T
3T
4T
5T
6T
A cos(2$ft)
-A cos(2$ft)
Example of Demodulation
A {1 + cos(4$fct)} -A {1 + cos(4$fct)}
+2A
After multiplication at receiver x(t) cos(2$fct)
0 T 2T 3T 4T 5T 6T
2A +A
0 T 2T 3T 4T 5T 6T
-A 1 0 1 1 0 1
Recovered Information
If Baseband signal x(t) with bandwidth B Hz then Modulated signal x(t)cos(2$fct) has bandwidth 2B Hz
!
f fc-B fc fc+B
Then baseband channel has Wc/2 Hz available, so modulation system supports Wc/2 x 2 = Wc pulses/second That is, Wc pulse/s per Wc Hz = 1 pulse/Hz Recall baseband transmission system supports 2 pulses/Hz
Ak
x
cos(2$fct)
Yi(t) = Ak cos(2$fct)
+
Yq(t) = Bk sin(2$fct)
Y(t)
Transmitted Signal
Bk
x
sin(2$fct)
! !
Yi(t) and Yq(t) both occupy the bandpass channel QAM sends 2 pulses/Hz
QAM Demodulation
Y(t)
x
2cos(2!fct)
Ak
x
2sin(2!fct)
Bk
Complex numbers
!
Tx signal with in-phase and quad modulation is x(t) = Akcos(2$fct)+Bksin(2$fct) Let Ck=Ak+iBk; then x(t) = Re[Ckexp(i2$fct)] We can now generalize to any alphabet of complex symbols to code information bits.
!
E.g. in case of polar coordinates: Re[Mkexp(i'k)exp(i2$fct)] = Mkcos(2$fct+'k) This is phase modulation ifMk is a constant or mixed amplitude and phase modulation in the general case (QAM)
Signal Constellations
! !
Each pair (Ak, Bk) defines a point in the plane Signal constellation set of signaling points
Bk
(-A,A) (A, A)
Bk
Ak
(-A,-A) (A,-A)
Ak
Bk
Bk
Ak
Ak
Fundamentals of communications
Properties of Media and Digital Transmission Systems
Adapted from slides of the book: A. Leon Garcia, I. Widjaja, Communication networks, McGraw Hill, 2004
Amplitude response & bandwidth Susceptibility to noise & interference c = 3 x 108 m/s in vacuum . = c/"/ speed of light in medium where />1 is the dielectric constant of the medium . = 2.3 x 108 m/s in copper wire; . = 2.0 x 108 m/s in optical fiber
Analog telephone
Frequency (Hz)
Optical fiber
102 104
106
108
Ultraviolet light
Infrared light
Visible light
106
104
102
10
Gamma rays
Microwave radio
Broadcast radio
X-rays
Attenuation
!
Dependence on distance of central importance Received power at d meters proportional to 100d Attenuation in dB = 0d, where 0 is dB/meter Received power at d meters proportional to d-n Attenuation in dB = n log d, where n is path loss exponent; n=2 in free space Signal level maintained for much longer distances Space communications possible
! !
Twisted Pair
Twisted pair ! Two insulated copper wires arranged in a regular spiral pattern to minimize interference ! Various thicknesses, e.g. 0.016 inch (24 gauge) ! Low cost ! Telephone subscriber loop from customer to CO ! Intra-building telephone from wiring closet to desktop ! LAN cabling (Fast Ethernet, GbE) 30 26 gauge 24 gauge 22 gauge 19 gauge 12 6 1 f (kHz) 10 100 1000 Higher attenuation rate for DSL
Attenuation (dB/mi)
24 18
1.544 18,000 feet, Mbps 5.5 km 6.312 12,000 feet, Mbps 3.7 km 12.960 Mbps 25.920 Mbps 51.840 Mbps 4500 feet, 1.4 km 3000 feet, 0.9 km 1000 feet, 300 m
!
Twisted pairs can provide high bit rates at short distances Asymmetric Digital Subscriber Loop (ADSL)
! !
Strategy for telephone companies is to bring fiber close to home & then twisted pair Higher-speed access + video
Ethernet LANs
! ! !
! ! ! ! ! !
Category 3 unshielded twisted pair (UTP): ordinary telephone wires Category 5 UTP: tighter twisting to improve signal quality Shielded twisted pair (STP): to minimize interference; costly 10BASE-T Ethernet ! 10 Mbps, Baseband, Twisted pair ! Two Cat3 pairs ! Manchester coding, 100 meters 100BASE-T4 Fast Ethernet ! 100 Mbps, Baseband, Twisted pair ! Four Cat3 pairs ! Three pairs for one direction at-a-time ! 100/3 Mbps per pair; ! 3B6T line code, 100 meters Cat5 & STP provide other options
Coaxial Cable
!
! !
! !
Cylindrical braided outer conductor surrounds insulated inner wire conductor High interference immunity Higher bandwidth than twisted pair ! Hundreds of MHz Cable TV distribution Long distance telephone transmission Original Ethernet LAN medium
35 30 0.7/2.9 mm 1.2/4.4 mm
Attenuation (dB/km)
25 20 15 10 5 0.1 1.0
2.6/9.5 mm
10
100 f (MHz)
42 MHz
500 MHz
750 MHz
550 MHz
! ! ! !
Cable TV network originally unidirectional Cable plant needs upgrade to bidirectional 1 analog TV channel is 6 MHz, can support very high data rates Cable Modem: shared upstream & downstream
! !
5-42 MHz upstream into network; 2 MHz channels; 500 kbps to 4 Mbps >550 MHz downstream from network; 6 MHz channels; 36 Mbps
5 MHz
Head end
54 MHz
Fiber node
Fiber
Fiber node
Fiber
Optical Fiber
Electrical signal Modulator Optical fiber Receiver Electrical signal
Optical source
!
Light sources (lasers, LEDs) generate pulses of light that are transmitted on optical fiber
! ! !
Very long distances (>1000 km) Very high speeds ( up ro 40 Gbps/wavelength) Nearly error-free (BER of 10-15) Dominates long distance transmission Distance less of a cost factor in communications Plentiful bandwidth for new services
Cladding Core
Jacket
1c
! ! !
Very fine glass cylindrical core surrounded by concentric layer of glass (cladding) Core has higher index of refraction than cladding Light rays incident at less than critical angle 1c is completely reflected back into the core
Rays on different modes interfere causing dispersion & limiting bit rate More expensive lasers, but achieves very high speeds
Disadvantages
!
Very low attenuation Immunity to external e.m interference Extremely high bandwidth Security: Very difficult to tap without breaking No corrosion More compact & lighter than copper wire Wideband optical amplifiers available (EDFA, SOA)
Shot noise Polarization dependence Non linear effects If physical arc of cable too high, light lost or wont reflect Will break
! !
! !
! !
Loss (dB/km)
0.8
1.0
1.2
1.4
1.6
1.8
Wavelength (m)
100 50 10
v 21+"2 ! v "2 21 2
Loss (dB/km)
"2/21 1 + "2/21
5 1 0.5 0.1
0.8
1.0
1.2
1.4
1.6
1.8
Different wavelengths carry separate signals Multiplex into shared optical fiber Each wavelength like a separate circuit A single fiber can carry 160 wavelengths, 10 Gbps per wavelength: 1.6 Tbps!
21 22 21 22 . 2m
21 22
2m
optical mux
optical fiber
optical demux
2m
1540
1550
1560
100 GHz
100 GHz
wavelength
band C
1540 1560 1580
band L
1600 1620
OA Erbium:Silica Erbium:
100 GHz
40
1440 1460 14801520
40
1540 1560 1580
40
1600 1620
80
1440 1460 14801520
80
1540 1560 1580
80
1600 1620 50 GHz
Attenuation Dispersion: chromatic (guide + material), polarization (PMD) Non linear effect: Brillouin, Raman, Kerr (Self-Phase Modulation (SPM), Cross-Phase Modulation (CPM), Four Wave Mixing (FWM))
!
Launched power of 20&mW (13 dBm) in the cross section of a monomodal fiber (50 'm2) corresponds to 40 kW/cm2.
! ! !
Crosstalk in (D)WDM systems Noise of optical amplifiers (ASE) Laser phase noise
Optical fiber response to an input pulse x(t) is the sum of two terms:
! !
Let x(t) be the sum of three sinusoidal terms with frequencies f1, f2, f3
!
a linear dispersive term yL(t) = "x(()h(t()d( an instantaneous cubic non linear term yNL(t) = Ax3(t)
Besides linear terms, at the output we can find sinusoidal signals at frequencies f1f2f3
!
OA
input
output
Effect of FWM
Power spectrum of an 8x10 Gbps DWDM signal with Unequal Channel Spacing after 50 km of G.653 fiber (2 mW/ch).
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
The maximum span of an optical signal is determined by the available power & the attenuation:
!
If 30 dB attenuation are allowed, then at 1550 nm, optical signal attenuates at 0.25 dB/km, so max span = 30 dB/0.25 km/dB = 120 km
Optical amplifiers increase optical signal power (no equalization, no regeneration): Pout=GPin+PASE.
! !
Noise is added by each amplifier SNRout < SNRin, so there is a limit to the number of OAs in an optical path
Requires optical-to-electrical (O-to-E) signal conversion, equalization, detection and retransmission (E-to-O) Expensive
! ! !
DWDM system carries many signals in one fiber At each span, a separate regenerator required per signal Very expensive
R R R R
DWDM multiplexer
R R
R R
R R
R R R
Optical Amplifiers
!
Optical amplifiers can amplify the composite DWDM signal without demuxing or O-to-E conversion Erbium Doped Fiber Amplifiers (EDFAs) boost DWDM signals within 1530 to 1620 range
! !
Spans between regeneration points >1000 km Number of regenerators can be reduced dramatically with dramatic reduction in cost of long-distance communications
OA
Optical amplifier
OA
R R R
OA
OA
R R R
Radio Transmission
! ! !
Radio signals: antenna transmits sinusoidal signal (carrier) that radiates in air/space Information embedded in carrier signal using modulation, e.g. QAM Communications without tethering
!
! ! !
Multipath propagation causes fading; iintercation of e.m. field with obstacles causes shadowing Interference from other users Spectrum regulated by national & international regulatory organizations
Radio Spectrum
Frequency (Hz) 104 105 106 107 108 109 1010 1011 1012
FM radio and TV AM radio Wireless cable Cellular and PCS Satellite and terrestrial microwave LF 10
4
MF 103
HF 102
VHF 101 1
UHF 10-1
SHF
Spectrum allocation
Short-Wave Radio FM Broadcast Audio Frequencies AM Broadcast Television Telecommunications PCS
Visible Light
Ultra High
Ultra Violet
Very High
Very Low
Medium
Extremely Low
High
Low
Microwave
Infrared
X-rays
Source: Motorola
2.4GHz
5.1GHz
2440
France 2480
2500
902 MHz ! 26 MHz BW ! Crowded ! Worldwide limited 2.4 GHz ! 83.5 MHz BW ! Available worldwide ! IEEE802.11 WLANs 5.1 GHz ! 300 MHz BW discontinuous ! Developing
U-NII Europe HiperLAN2* 5700 5800 5900
5100
5400
5500
5600
Examples
Cellular Phone ! Allocated (licensed) spectrum ! First generation: ! 800, 900 MHz ! Initially analog voice ! Second generation: ! 1800-1900 MHz ! Digital voice, messaging Wireless LAN ! Unlicenced ISM spectrum ! Industrial, Scientific, Medical ! 902-928 MHz, 2.400-2.4835 GHz, 5.725-5.850 GHz ! IEEE 802.11 LAN standard ! 11-54 Mbps Point-to-Multipoint Systems ! Directional antennas at microwave frequencies ! High-speed digital communications between sites ! High-speed Internet Access Radio backbone links for rural areas Satellite Communications ! Geostationary satellite @ 36000 km above equator ! Relays microwave signals from uplink frequency to downlink frequency ! Long distance telephone ! Satellite TV broadcast
1920-1980 for Uplink.(12x5MHz) 2110-2170 for Downlink.(12x5Mhz) 1900-1920 (4x5Mhz) 2010-2025 (3x5Mhz)
Current 3G scenario
UMTS Operator
UTRAN Node B RNC UMTS IPv4 Core Network SGSN Node B GGSN IPv4 backbone
Signaling Gateway
WWW E-mail .,
Profiles DBs ,
UMTS: (3GPP) MM: Ad-Hoc Protocols AA: Ad-Hoc Protocols WLAN: (IETF) MM: Mobile IP, Cellular IP, etc AA: User Name/Pwd
Applications
! ! ! !
Local wireless networking (infrastructured, ad hoc) Public hot spots Home Networking Sensor networks
Fundamentals of communications
Error Detection and Correction
Adapted from slides of the book: A. Leon Garcia, I. Widjaja, Communication networks, McGraw Hill, 2004
Error Control
! !
Digital transmission systems introduce errors Applications require certain reliability level
! !
Data applications require error-free transfer Voice & video applications tolerate some errors, the less the more source coding removes redundancy from original signal
Error control used when transmission system does not meet application requirement Two basic approaches: Error detection & retransmission (ARQ) Forward error correction (FEC)
Key Idea
!
If received block doesnt satisfy pattern, it is in error If it satisfies pattern, it is assumed to be correct
Example: spell checking by taking dictionary words with same initial as letter to be spelled out
All inputs to channel satisfy pattern or condition User Encoder information Channel
Channel output Pattern checking Deliver user information or set error alarm
Given a data block of k bits, add one more bit so as to make the number of 1s even
!
b1, b2, b3, , bk bk+1 = b13b23b33 3bk (b1, b2, b3, , bk,, bk+1)
All error patterns that change an odd # of bits are detectable; all even-numbered patterns are undetectable
Information (7 bits): (0, 1, 0, 1, 1, 0, 0) Parity Bit: b8 = 0 + 1 +0 + 1 +1 + 0 = 1 Codeword (8 bits): (0, 1, 0, 1, 1, 0, 0, 1) If single error in bit 3 : (0, 1, 1, 1, 1, 0, 0, 1)
! !
k bit
append nk bit
Sent codeword
n bit Channel
Received codeword
n bit
Received check bits Information accepted if check bits match
Compare
Redundancy: Single parity check code adds 1 redundant bit per k information bits: overhead = 1/(k+1) Coverage: all error patterns with odd # of errors can be detected
!
An error patten is a binary (k + 1)-tuple with 1s where errors occur and 0s elsewhere Of 2k+1 binary (k+1)-tuples, 1/2 have odd weight, so 50% of error patterns can be detected
Is it possible to detect more errors if we add more check bits? Yes, with the right codes
Many transmission channels introduce bit errors at random, independently of each other, with probability p Some error patterns are more probable than others: P[10000000] = p(1 p)7 = (1 p)8 P[11000000] = p2(1 p)6 p 1p p)8 and p 2 1p
= (1
It follows that patterns with 1 error are more likely than patterns with 2 errors and so forth
P[undetectable error] =
1 0 0 1 0 0 0 1 0 0 0 1 Last column consists 1 0 0 1 0 0 of check bits for each 1 1 0 1 1 0 row 1 0 0 1 1 1 Bottom row consists of check bit for each column
Error-detecting capability
1 0 0 1 0 0 0 0 0 0 0 1 1 0 0 1 0 0 1 1 0 1 1 0 1 0 0 1 1 1 One error 1 0 0 1 0 0 0 0 0 0 0 1 Two 1 0 0 1 0 0 errors 1 0 0 1 1 0 1 0 0 1 1 1
1 0 0 1 0 0 0 0 0 1 0 1 1 0 0 1 0 0 Three errors 1 0 0 1 1 0 1 0 0 1 1 1
1 0 0 1 0 0 0 0 0 1 0 1 1 0 0 1 0 0 1 0 0 0 1 0 1 0 0 1 1 1 bits
1, 2, or 3 errors can always be detected; Not all patterns >4 errors can be detected
Four errors
(undetectable)
Many channels have preference for error patterns that have fewer # of errors These error patterns map transmitted codeword to nearby n-tuple If codewords close to each other then detection failures will occur Good codes should maximize separation between codewords
o o o o x x x x x o o o x x o o o o o
x = codewords o = noncodewords
o x x o o
o o x
x o
o o x
o o o x o x
Single parity check codes do not detect enough errors Two-dimensional codes require too many check bits
The following error detecting codes are used in practice: ! Internet Check Sums ! CRC Polynomial Codes
Internet Checksum
!
Several Internet protocols (e.g. IP, TCP, UDP) use check bits to detect errors in the IP header (or in the header and data for TCP/UDP)
!
A checksum is calculated for header/segment contents and included in a special field. Checksum recalculated at every router, so algorithm selected for ease of implementation in software
Let header consist of L, 16-bit words, b0, b1, b2, ..., bL1 The algorithm appends a 16-bit checksum bL
Checksum Calculation
The checksum bL is calculated as follows:
!
Treating each 16-bit word as an integer, find x = b0 + b1 + b2 + ... + bL1 modulo 2161 The checksum is then given by: bL = x modulo 2161 Thus, the headers must satisfy the following pattern: 0 = b0 + b1 + b2+ ...+ bL1 + bL modulo 2161
The checksum calculation can be carried out in software at speed compatible with router operations
Polynomial Codes
!
Coefficients of polynominal are orderly equal to bits of the codeword, e.g. MSB being the coefficient of the leading power of the polynomial
! ! ! !
Polynomial arithmetic instead of check sums Implemented using shift-register circuits Also called cyclic redundancy check (CRC) codes Most data communications standards use polynomial codes for error detection
!
Binary vectors map to polynomials (bk-1 , bk-2 ,, b2 , b1 , b0) # bk-1xk-1 + bk-2xk-2 + + b2x2 + b1x + b0
Addition:
(x7 + x6 + 1) + (x6 + x5) = x7 + x6 + x6 + x5 + 1 = x7 +(1+1)x6 + x5 + 1 = x7 +x5 + 1 since 1+1=0 mod2
Multiplication:
(x + 1) (x2 + x + 1) = x(x2 + x + 1) + 1(x2 + x + 1) = x3 + x2 + x) + (x2 + x + 1) = x3 + 1
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi
Polynomial Division
divisor
Polynomial Coding
!
Code has binary generating polynomial of degree nk g(x) = xn-k + gn-k-1xn-k-1 + + g2x2 + g1x + 1 k information bits define polynomial of degree (k1 i(x) = ik-1xk-1 + ik-2xk-2 + + i2x2 + i1x + i0 Find remainder polynomial of at most degree nk1 q(x) xn-ki(x) = q(x)g(x) + r(x) g(x) | xn-k i(x) r(x) Define the codeword polynomial of degree (n1
x5 + x4 + x3
b = (1,1,0,0,0,1,0)
G(x) = x 3 + x +1
g1 = 1
M(x) = x 3 + x
g0 = 1
g3 = 1
Reg 1 Reg 2
M(x)
!
Reg 0
!
!
All codewords are a multiple of g(x)! Receiver should divide received n-tuple by g(x) and check if remainder is zero If remainder is nonzero, then received n-tuple is not a codeword
b(x)
R(x)=b(x)+e(x)
! !
e(x) has 1s in error locations & 0s elsewhere Receiver divides the received polynomial R(x) by g(x) Blindspot: If e(x) is a multiple of g(x), that is, e(x) is a nonzero codeword, then R(x) = b(x) + e(x) = q(x)g(x) + q(x)g(x) The set of undetectable error polynomials is the set of nonzero code polynomials Choose the generator polynomial so that most common error patterns can be detected.
Select generator polynomial so that likely error patterns are not multiples of g(x)
For e(x) to be an undetectable error pattern, it must be e(x) = q(x)g(x) for some q(x) Evaluate this identity at x=1 and get e(1) = 0 (it must be g(1)=0; why?); then e(x) cannot have ad odd number of 1s (why?)
!
CCITT-32: = x32 + x26 + x23 + x22 + x16 + x12 + x11 + x10 + x8 + x7 + x5 + x4 + x2 + x + 1 IEEE 802, DoD, V.42
Hamming Codes
! ! !
Class of error-correcting codes Capable of correcting all single-error patterns For each m > 2, there is a Hamming code of length n = 2m 1 with n k = m parity check bits
Redundancy
m
3 4 5 6
n = 2m1
7 15 31 63
k = nm
4 11 26 57
m/n
3/7 4/15 5/31 6/63
m = 3 Hamming Code
! !
Information bits are b1, b2, b3, b4 Equations for parity checks b5, b6, b7 b5 = b1 b6 = b1 + b2 b7 = + b3 + b4 + b4 + b 2 + b3 + b4
! !
b1
0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1
b2
0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1
b3
0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1
b4
0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
b1 b2 b3 b4 b5 b6 b7
0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 0 0 1 1 0 1 0 0 1 1 0 0 1 0 1 0 1 1 0 1 0 1 0 1 0 0 1 0 1 0 1 1 0 1 0 0 1 0 1 1 0 1 0 0 1
w( b )
0 4 3 3 3 3 4 4 3 3 4 4 4 4 3 7
In matrix form: b1
b2
0 0 0 = 1011100 = 1101010 = 0111001
b3 b4 = H bt = 0 b5 b6 b7
!
All codewords must satisfy these equations Note: each nonzero 3tuple appears once as a column in check matrix H
= 0
1
s=He= 1101010
0111001
1011100
= 1 + 0
1 0
= 1
1
1 1 1 0 1011100 1 1 0 = 1 + 1 + 0 s=He= 1101010 0 1 0111001 0 1 0 Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi 0
= 0
With Hamming (7,4) code undetectable error pattern must have 3 or more bits, i.e. at least 3 bits must be changed to convert one codeword into another codeword
o o
b1
o
Distance 3 o o
b2
o
! !
Spheres of distance 1 around each codeword do not overlap If a single error occurs, the resulting n-tuple will be in a unique sphere around the original codeword
For m > 2, the Hamming code is obtained through the check matrix H:
! !
Each nonzero m-tuple appears once as a column of H The resulting code corrects all single errors
For each value of m, there is a polynomial code with g(x) of degree m that is equivalent to a Hamming code and corrects all single errors
!
R (Receiver)
e Error pattern
!
! !
The receiver first calculates the syndrome: s = HR = H (b + e) = Hb + He = He If s = 0, then the receiver accepts R as the transmitted codeword If s is nonzero, then an error is detected
! ! !
Hamming decoder assumes a single error has occurred Each single-bit error pattern has a unique syndrome The receiver matches the syndrome to a single-bit error pattern and corrects the appropriate bit
! ! !
Fundamentals of communications
RS-232 Asynchronous Data Transmission
Adapted from slides of the book: A. Leon Garcia, I. Widjaja, Communication networks, McGraw Hill, 2004
Serial line interface between computer and modem or similar device Data Terminal Equipment (DTE): computer Data Communications Equipment (DCE): modem Mechanical and Electrical specification
(a)
14 25
(b)
1 2 3 4 5 6 7 8 20 22
Protective Ground (PGND) Transmit Data (TXD) Receive Data (RXD) Request to Send (RTS) Clear to Send (CTS) Data Set Ready (DSR) Ground (G) Carrier Detect (CD) Data Terminal Ready (DTR) Ring Indicator (RI)
1 2 3 4 5 6 7 8 20 22
DTE
DCE
Synchronization
!
Data T S
1 0 1 1 1
Clock
0 0 1 0 0 0
Data T S Clock
Synchronization (contd)
! !
Asynchronous Transmission
!
Avoids synchronization loss by specifying a short maximum length for the bit sequences and resetting the clock in the beginning of each bit sequence. Accuracy of the clock?
3T/2
Synchronous Transmission
!
! ! !
R transition for R bits per second transmission R transition contains a sine wave with R Hz. R Hz sine wave is used to synch receiver clock to the transmitters clock using PLL (phase-lock loop)
1 Voltage
time
Telecomunicazioni - a.a. 2010/2011 - Prof. Andrea Baiocchi