dr. uri mahlab1. 2 communication system dr. uri mahlab3 + channel noise channel noise block diagram...
Post on 22-Dec-2015
221 views
TRANSCRIPT
Dr. Uri Mahlab 1
Dr. Uri Mahlab 2
Communication system
Dr. Uri Mahlab 3
++ Channel noiseChannel noise
Block diagram of an Binary/M-ary signalingBlock diagram of an Binary/M-ary signaling schemescheme
++
Dr. Uri Mahlab 4
Block diagram DescriptionBlock diagram Description
{d{dkk}={1,1,1,1,0,0,1,1,0,0,0,1,1,1}}={1,1,1,1,0,0,1,1,0,0,0,1,1,1}
TTbb
TTbb
ForFor
ForFor
Dr. Uri Mahlab 5
Block diagram Description Block diagram Description (Continue - 1)(Continue - 1)
{d{dkk}={1,1,1,1,0,0,1,1,0,0,0,1,1,1}}={1,1,1,1,0,0,1,1,0,0,0,1,1,1}
Dr. Uri Mahlab 6
Block diagram Description Block diagram Description (Continue - 2)(Continue - 2)
{d{dkk}={1,1,1,1,0,0,1,1,0,0,0,1,1,1}}={1,1,1,1,0,0,1,1,0,0,0,1,1,1}
100110100110
Dr. Uri Mahlab 7
TimingTiming
Block diagram Description Block diagram Description (Continue - 3)(Continue - 3)
{d{dkk}={1,1,1,1,0,0,1,1,0,0,0,1,1,1}}={1,1,1,1,0,0,1,1,0,0,0,1,1,1}
TTbb
100110100110
tt
tt
Dr. Uri Mahlab 8
InformationInformation
sourcesourcePulsePulse
generatorgeneratorTransTransfilterfilter
) ) X(tX(tTimingTiming
Block diagram Description Block diagram Description (Continue - 4)(Continue - 4)
HHTT(f)(f)
TTbb 2T2Tbb
3T3Tbb 4T4Tbb
5T5Tbb
tt6T6Tbb
Channel noise n(t)Channel noise n(t)
++
tt
ReceiverReceiverfilterfilter
HHRR(f)(f)
tt
Dr. Uri Mahlab 9
Block diagram Description Block diagram Description (Continue - 5)(Continue - 5)
TTbb 2T2Tbb
3T3Tbb 4T4Tbb
5T5Tbb
tt6T6Tbb
tt
tt
Dr. Uri Mahlab 10
TimingTiming
++
Block diagram of an Binary/M-ary signalingBlock diagram of an Binary/M-ary signaling schemescheme
++
Dr. Uri Mahlab 11
Block diagram DescriptionBlock diagram Description
TTbb 2T2Tbb
3T3Tbb 4T4Tbb
5T5Tbb
tt6T6Tbb
tt
1 0 0 0 1 01 0 0 0 1 0
1 0 0 1 1 01 0 0 1 1 0
tt
tt
Dr. Uri Mahlab 12
Typical waveforms in a binary PAM system
Dr. Uri Mahlab 13
Block diagram of an Binary/M-ary signalingBlock diagram of an Binary/M-ary signaling schemescheme
TimingTiming
++
++
Dr. Uri Mahlab 14
Explanation of PExplanation of Prr(t)(t)
Dr. Uri Mahlab 15
The element of a baseband binary PAM system
Dr. Uri Mahlab 16
Dr. Uri Mahlab 17
For and is the total time delay in the For and is the total time delay in the system, we get. system, we get.
tt
tt
tt
tmY
tt11
tt22 tt33
ttmm
The input to the A/D converter isThe input to the A/D converter is
Dr. Uri Mahlab 18tt
tmY
tt11
tt22 tt33
ttmm
The output of the A/D converter at the sampling timeThe output of the A/D converter at the sampling time
ttm m =mT=mTbb+t+tdd
TTbb 2T2Tbb
3T3Tbb 4T4Tbb
5T5Tbb
tt6T6Tbb
Dr. Uri Mahlab 19tt
tmY
tt11
tt22 tt33
ttmm
ISI - ISI - IInter nter SSymbolymbolIInterferencenterference
Dr. Uri Mahlab 20
Explanation of ISIExplanation of ISI
tt
ff ff
tt
Dr. Uri Mahlab 21
Explanation of ISI - ContinueExplanation of ISI - Continue
tt
ff ff
tt
TTbb 2T2Tbb
3T3Tbb 4T4Tbb
5T5Tbb
tt6T6Tbb
Dr. Uri Mahlab 22
Dr. Uri Mahlab 23
-The pulse generator output is a pulse waveform
If kth input bit is 1
if kth input bit is 0
-The A/D converter input Y(t)
Dr. Uri Mahlab 24
Dr. Uri Mahlab 25
5.2 BASEBAND BINARY PAM SYSTEMS
- minimize the combined effects of inter symbol interference and noise in order to achieve minimum probability of error for given data rate.
Dr. Uri Mahlab 26
5.2.1 Baseband pulse shaping
The ISI can be eliminated by proper choice of received pulse shape pr (t).
Doe’s not Uniquely Specify Pr(t) for all values of t.
Dr. Uri Mahlab 27
TheoremTheorem
ProofProof
To meet the constraint, Fourier Transform Pr(f) of Pr(t), shouldTo meet the constraint, Fourier Transform Pr(f) of Pr(t), shouldsatisfy a simple condition given by the following theoremsatisfy a simple condition given by the following theorem
Dr. Uri Mahlab 28
b
b
T2/1
T2/1
bbbr df)fnT2jexp(T)nT(p
Which verify that the Pr(t) with a transform Pr(f) Which verify that the Pr(t) with a transform Pr(f) Satisfy Satisfy ____________________________
Dr. Uri Mahlab 29
The condition for removal of ISI given in the theorem is calledThe condition for removal of ISI given in the theorem is calledNyquist (Pulse Shaping) CriterionNyquist (Pulse Shaping) Criterion
TTbb 2T2Tbb-T-Tbb-2T-2Tbb
11
Dr. Uri Mahlab 30
The Theorem gives a condition for the removal of ISI using a Pr(f) withThe Theorem gives a condition for the removal of ISI using a Pr(f) witha bandwidth larger then rb/2/. a bandwidth larger then rb/2/.
ISI can’t be removed if the bandwidth of Pr(f) is less then rb/2.ISI can’t be removed if the bandwidth of Pr(f) is less then rb/2.
TTbb 2T2Tbb
3T3Tbb 4T4Tbb
5T5Tbb
tt6T6Tbb
Dr. Uri Mahlab 31
pr )t(
Particular choice of Pr(t) for a given application
Dr. Uri Mahlab 32
A Pr(f) with a smooth roll - off characteristics is preferable A Pr(f) with a smooth roll - off characteristics is preferable over one with arbitrarily sharp cut off characteristics.over one with arbitrarily sharp cut off characteristics.
Pr(f)Pr(f) Pr(f)Pr(f)
Dr. Uri Mahlab 33
In practical systems where the bandwidth available for In practical systems where the bandwidth available for transmitting data at a rate of rtransmitting data at a rate of rbb bits\sec is between r bits\sec is between rbb\2 to r\2 to rbb Hz, a class of pHz, a class of prr(t) with a (t) with a raised cosine frequency raised cosine frequency characteristiccharacteristic is most commonly used.is most commonly used.A raise Cosine Frequency spectrum consist of a flat amplitude portion and a roll off A raise Cosine Frequency spectrum consist of a flat amplitude portion and a roll off portion that has a sinusoidal form.portion that has a sinusoidal form.
Dr. Uri Mahlab 34
raised cosine frequency characteristicraised cosine frequency characteristic
Dr. Uri Mahlab 35
The BW occupied by the pulse spectrum is The BW occupied by the pulse spectrum is B=rB=rbb/2+/2+. . The minimum value of B is rb/2 and the maximum value is The minimum value of B is rb/2 and the maximum value is rrbb..
Larger values ofLarger values of imply that moreimply that more bandwidth is required for a bandwidth is required for a given bit rate, however it lead for faster decaying pulses, whichgiven bit rate, however it lead for faster decaying pulses, whichmeans that synchronization will be less critical and will not cause means that synchronization will be less critical and will not cause large ISI.large ISI.
=rb/2 leads to a pulse shape with two convenient properties. =rb/2 leads to a pulse shape with two convenient properties. The half amplitude pulse width is equal to Tb, and there are zero The half amplitude pulse width is equal to Tb, and there are zero crossings at t=3/2Tb, 5/2Tb…. In addition to the zero crossingcrossings at t=3/2Tb, 5/2Tb…. In addition to the zero crossing at Tb, 2Tb, 3Tb,…...at Tb, 2Tb, 3Tb,…...
SummarySummary
Dr. Uri Mahlab 36
5.2.25.2.2
Optimum transmitting and receiving Optimum transmitting and receiving filtersfilters
The transmitting and receiving filters are chosen to provideThe transmitting and receiving filters are chosen to providea propera proper
Dr. Uri Mahlab 37
-One of design constraints that we have for selecting the filters -One of design constraints that we have for selecting the filters is the relationship between the Fourier transform of pis the relationship between the Fourier transform of prr(t) and (t) and ppgg(t).(t).
In order to design optimum filter Ht(f) & Hr(f), we will assume that Pr(f), In order to design optimum filter Ht(f) & Hr(f), we will assume that Pr(f), Hc(f) and Pg(f) are known. Hc(f) and Pg(f) are known.
Where tWhere tdd, is the time delay Kc normalizing constant., is the time delay Kc normalizing constant.
Portion of a baseband PAM systemPortion of a baseband PAM system
Dr. Uri Mahlab 38
If we choose Pr(t) {Pr(f)} to produce Zero ISI we are left If we choose Pr(t) {Pr(f)} to produce Zero ISI we are left only to be concerned with noise immunity, that is will chooseonly to be concerned with noise immunity, that is will choose
Dr. Uri Mahlab 39
Noise ImmunityNoise Immunity
Problem definition:Problem definition:
For a givenFor a given : :•Data Rate - Data Rate - •Transmission power -Transmission power -•Noise power Spectral Density -Noise power Spectral Density -•Channel transfer function -Channel transfer function -•Raised cosine pulse -Raised cosine pulse -
Choose Choose
Dr. Uri Mahlab 40
Error probability CalculationsError probability Calculations
At the m-th sampling time the input to the A/D is:At the m-th sampling time the input to the A/D is:
We decideWe decide::
Dr. Uri Mahlab 41
A=aKcA=aKc
The noise is assumed to be zero mean Gaussian at the receiver inputThe noise is assumed to be zero mean Gaussian at the receiver inputthen the output should also be Zero mean Gaussian with variance Nothen the output should also be Zero mean Gaussian with variance No
given by:given by:
Dr. Uri Mahlab 42bb
Dr. Uri Mahlab 43
00
Dr. Uri Mahlab 44
-A-A AA
Dr. Uri Mahlab 45
Dr. Uri Mahlab 46
UU
Q(u)Q(u)
dz=dz=
Dr. Uri Mahlab 47
0N
A
PPerrorerror decreases as decreases as increaseincrease
Hence we need to maximize the signalHence we need to maximize the signalto noise Ratioto noise Ratio
Thus for maximum noise immunity the filter transfer functions HThus for maximum noise immunity the filter transfer functions HTT(f)(f)and Hand HRR(f) must be xhosen to maximize the SNR(f) must be xhosen to maximize the SNR
Dr. Uri Mahlab 48
Optimum filters design calculationsOptimum filters design calculations
We will express the SNR in terms of HWe will express the SNR in terms of HTT(f) and H(f) and HRR(f) (f)
We will start with the signal:We will start with the signal:
The PSD of the transmitted signal is given by::The PSD of the transmitted signal is given by::
Dr. Uri Mahlab 49
And the average transmitted power ST isAnd the average transmitted power ST is
The average output noise power of nThe average output noise power of n00(t) is given by:(t) is given by:
Dr. Uri Mahlab 50
The SNR we need to maximize is The SNR we need to maximize is
Or we need to minimize Or we need to minimize
Dr. Uri Mahlab 51
Using Schwartz’s inequality Using Schwartz’s inequality
The minimum of the left side equaity is reached whenThe minimum of the left side equaity is reached when
V(f)=Const*W(f)V(f)=Const*W(f)
If we choose :If we choose :
Dr. Uri Mahlab 52
whenminimized is 2
The filter should have a linear phase response in a total time delay of tdThe filter should have a linear phase response in a total time delay of td
Dr. Uri Mahlab 53
Finally we obtain the maximum value of the SNR to be:Finally we obtain the maximum value of the SNR to be:
Dr. Uri Mahlab 54
For AWGN with For AWGN with
andand
ppgg(f) is chosen such that it does not change much over the (f) is chosen such that it does not change much over the bandwidth of interest we get. bandwidth of interest we get.
Rectangular pulse can be used at the input of HRectangular pulse can be used at the input of HTT(f).(f).
Dr. Uri Mahlab 55
5.2.3 Design procedure and Example
The steps involved in the design procedure.
Example:Design a binary baseband PAM system to transmit data at a a bit rate of 3600 bits/sec with a bit error probability less than .10 4
The channel response is given by:
The noise spectral density is HzwattfGn /10)( 14
Dr. Uri Mahlab 56
Solution:
HzwattfG
HzB
p
bitsr
n
e
b
/10)(
2400
10
sec/3600
4
4
If we choose a braised cosine pulse spectrum with6006/ br
Dr. Uri Mahlab 57
We choose a pg(t)
We choose
)()()()()(
)10)(3600( 31
fpfHfHfHfp
K
rRcTg
Dr. Uri Mahlab 58
Plots of Pg(f),Hc(f),HT(f),HR(f),and Pr(f).
Dr. Uri Mahlab 59
To maintain a 410eP
For Pr(f) with raised cosine shape
1)( dffPr
And hence dBmST 23)10)(3600)(06.14( 10
Which completes the design.
Dr. Uri Mahlab 60
In order to transmit data at a rate of rb bits/sec with zero ISI PAM data transmission system requires a bandwidth of at least rb
/2 HZ.
+Binary PAM data transmission at a rate of rb bits/sec with zero ISI
1.such filters are physically unrealizable
2. Any system with this filters will be extremely sensitive to perturbations.
204
5.3
Dr. Uri Mahlab 61
Dr. Uri Mahlab 62
The duobinary signaling schemes use pulse spectra Pr(f):
The duobinary scheme utilizes controlled amounts of ISI for transmitting data at a rate of rb /2 HZ. The shaping filters for duobinary are easier to realize than the ideal rectangular filters.
The pulse response Pr(t): tPr
Dr. Uri Mahlab 63
The output Y(t) of the receive filter can be written as:
The Am' s can assume one of two values +/-A depending on whether the m th input bit is 1 or 0. Since Y(tm) depends on Am & Am-1 assuming no noise :
+2A if the m th and (m-1st)bits are both 1's
0 if the m th and (m-1st)bits are different
-2A if the m th and (m-1st)bits are both zero
If the output is sampled at tm= mTb/2+ td than it is obvious that in the absence of noise:
Y(tm) =
5.3.1
Dr. Uri Mahlab 64
Transmitting and receiving filters for optimum performance
Since no is a zero mean Gaussian random variable with a
variance No we can write pe as:
The receiving levels at the input to the A/D converter are 2A, 0, and -2A with probabilities 1/2, 1/4. The probability of bit error pe is given by:
For the direct binary PAM case :
5.3.2
Dr. Uri Mahlab 65
The integral can be evaluated as :
Where /2=Gn(f) is the noise power spectral density the probability of error is:
Dr. Uri Mahlab 66
M-ARY SIGNALING SCHEMES
In baseband binary PAM we use pulses with one of 2 possible amplitude, In M-ary baseband PAM system we allowed M possible levels (M>2) and there M distinct input symbols.
ss rT 1During each signaling interval of duration the source is converted to a four-level PAM pulse train by the the pulse generator.
The signal pulse noise passes through the receiving filter and is sampled by the A/D converter at an appropriate rate and phase.
the M-ary PAM scheme operating with the preceding constraints can transmit data at a bit rate of rs log2M bit/sec and require a minimum bandwidth of rs/2 HZ.
5.4
Dr. Uri Mahlab 67
5.7 MISCELLANEOUS TOPICS
5.7.1 Eye Diagram The performance of baseband PAM systems depends on the amount of ISI and channel noise.
The received waveform with no noise and no distortion is shown in Figure 5.20a the “open” eye pattern results
Figure 5.20b shows a distorted version of the waveform the corresponding eye pattern.
Figure 5.20c shows a noise distorted version of the received waveform and the corresponding eye pattern.
Dr. Uri Mahlab 68
Dr. Uri Mahlab 69
In the figure 5.21 we see typical eye patterns of a duobinary signal
Dr. Uri Mahlab 70
If the signal-to-noise ratio at the receiver is high then the following observations can be made from the eye pattern shown simplified in Figure 5.22:
Dr. Uri Mahlab 71
1. The best time to sample the received waveform is when the eye is opening is largest.
2. The maximum distortion is indicated by the vertical width of the two branches at sampling time.
3. The noise margin or immunity to noise is proportional to the width of the eye opening.
4. The sensitivity of the system to timing errors is revealed by the rate of closing of the eye as sampling time is varied.
5. The sampling time is midway between zero crossing.
6. Asymmetries in the eye pattern indicate nonlinearities in the channel.
Dr. Uri Mahlab 72
5.7.2 synchronization
Three methods in which this synchronization can be obtained are:
1. Derivation of clock information from a primary or secondary standard.
2. Transmitting a synchronizing clock signal.
3. Derivation of the clock signal from the received waveform itself.
An example of a system used to derive a clock signal from the received waveform is shown in figure 5.23.
To illustrate the operating of the phase comparator network let us look at the timing diagram shown in figure 5.23b
Dr. Uri Mahlab 73
Dr. Uri Mahlab 74
5.7.3 Scrambler and unscrambler
Scrambler: The scrambler shown in figure 5.24a consists a “feedback” shift register.
Unscrambler:The matching unscrambler have a ”feed forward” shift register structure.
In both the scrambler and unscrambler the outputs of several stages of of shift register are added together modulo-2 and the added to the data stream again in modulo-2 arithmetic
Dr. Uri Mahlab 75
Dr. Uri Mahlab 76
Scrambler affects the error performance of the communication system in that a signal channel error may cause multiple error at the output of the unscrambler.
The error propagation effect lasts over only a finite and small number of bits.
In each isolated error bit causes three errors in the final output it must also be pointed out that some random bit patterns might be scrambled to the errors or all ones.