0% found this document useful (0 votes)
21 views29 pagesDC Unit-3
I want to download the pdf of the day of the day of the day of
Uploaded by
sandeepsandy24240Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
21 views29 pagesDC Unit-3
I want to download the pdf of the day of the day of the day of
Uploaded by
sandeepsandy24240Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
/ 29
—. | [LEARNING QauecTives
Baseband receiver and its bit error probability
Ei cetinum fer and ts bw error probabity :
‘The matched filter and its bit error probability —*
of ASK, BPSK, BFSK and QPSK.
INTROQUCTION
its in the form of 1's and
can be represented in various ways lke, symbol 1 can be represented by @ voltage V and the
1° -¥ or by signals ,() ands, respective
‘The resulting signal can be transmitted directly without
sion of data over the channel, he signal will be corrupted by noise and hence there will be
iver will make an error while determining the bit ie, oF O was transmitted. In this unit, we
jes arid methods to minimize them. Further we will study the concepts of matched fil
‘Scanned with CamSeanner ba
Q1. With a noat sketch oxplain the
. at 8 (1)
as itustyaed in figure (
A baseband signal receiver
Figure (1): Basobond signal Recolv
oro
mn cera
i Seon a Meo
imine a eat ee ane
ete:
Pied Ld epi
pie etn i ang
(0),
SM Wn SL
Pe Abs
rom figure (2) it can be obsefved th
RE ae eats kt et
ce 4 va
Me nn a
lar wave
Ea Ey
PLS Ste nov rent nites
Mee ee eet
USS eee
expression for signal to RMS nol cP eee
eer
TL ea ree
Cast = ~@)
signal power in standard 10.resistance is,
er
: transfer function of an integrator for one bit duration
1-9 _ 1-[coso? ~ jsinw7]
“jor jor
_ Lecosw? + jsinoT
ie jor
inoT | 1—coswf.
ae
‘of +1+cos*oT —2.cosoT
ot
sensing raion in) we ae .
SO = we 2
Substituting equation (10) in (7) we Bet
m0 - funy
sin’@7!2 1
= 34
a (fy? y
(From equation (6)
Let, aft =x 7
de = nedf
Pye
Es
nd
r= 7 (3)
Substituting the — equation (11), we ge
ay a"
‘Scanned with CamSeanner DIGITAL COMMUNICATIONS [JNTU- KAK Nag,
Error usinc Matcuep
eo arr
oF
PROBABILITY
3.2 THe Orrimum Ficter, MatcHeD FILTER,
iter.
peer meee ag
Co
ed fA
Draw and explain the receiver for a binary coded sig}
rnd
yaa Pee
x Teen eee cena ree nee see et
‘generalized filter for receiving binary
OU Tt ad Sea eee
Fel Resa ewes arc a
ed signal is defined as a series of electrical pi , 5 Seales
Sores e as pee een ag me ie ST
Neeag it Berman 7 any impact on the functioning Of the receiver. Thus, 1 ae a
DRCNet ect cs 2) Seah
i s paren als x,(#) and x,(2), The main purpose of this receiyer is to minimize the pr
HOeun sn ere ee ‘
Ried TROae ULE aon estan ecea ese to Tem ere ra oa
x. The basic block diagram «
n,(T)
)+n(T)
iss
Oeste Onna)
n() ~ Noise added over the channel,
A receiver of binary coded signal consists of an optimum, ee Tee
oem nee TT acct ee) SS alae are
Tesh aae tian a is
eee
OK enrte
receiver, which idee y
Ue DETTE
Seat rey se
MCA aaa tg Tir Ta
AO Meme atd ta}
Then, the input to the optimum filter is given by,
erie tg)
The output of the receiver with noise is expresed as,
FD) =s,(1) + n(D)
or
AURA Ee TG
Me Aaa ati reed
eres
AUCs
when, s(1) = » (7)
FAT) =$(1), when, s(1)= » (4
We RU Sean ee
Gagan
btu conn ane
ETO een
TRC Ta Aco irre ta Ta aang AUR (ccraty
lies between s,.(7) and %n(T) and is expressed as,
So(T)+ 599(T)
Sauna ae eae
Thus, the decision boundary
asa sil
Decision boundary, V, ~
ed with Cam n(D2V 5D)
ninz 1D * aD
ul eee
n(n 22a
= 2
=
f = @)
‘Then the probability of error is,
pee [cones
BP lien nse RE ee
2
3
las i es
sag
i ere
a
i
"Substituting the above relations in equation (3), we get,
(01
Calculate the transfer function
‘optimum filter.
Ans:
‘To determine the minimum error pro
the input signal to the optimum filter be p() gt
PO) =5(0 3,0. +
‘The corresponding output of the filter is
= p= 80-1 .
Let P(f) and P.(f) be the Fourier transforms of p() and
If HG) is the transfer function of the filter then
PD =HOPO | hid
mndpi=f Dea,
‘Substituting equation (3) in (4), we Bet,
pin = | HOrGe™ a On
‘The input noise to the optimum filter is n(@) and the
corresponding output nose isn, that has a power spectral
density G_(f. The psd of output noise and the psd of input
noise, G,(f)is related as, ‘
6.) = HOP GD 6
‘The normalized output noise power ic, the noise
variance, o? is related to the output noise psd of the filter as,
fone
6!
= a= flHprena
Consider a parameter, y given
y= falD=taD
Scanned with CamScanner ‘Substituting equations (5) and (7) in (9) we get,
ele af | Lemnnral
(19) | funetion HY).
substituting equations (13) and (14) in (12) the
[ures Cee cate
‘The schwartz inequality states that given arbitrary
complex functions X(/) and ¥() of a common variable f then
| Fone « Jixoriar frovy
\ (or)
if Fixoparsinnra 1) =
iL ora [ro tee Ces ace na auc ree hres
“ OL a eee
‘and equal sign applies when
XQ) = KY) (12)
Where,
k= Arbitrary constant
YO) ~Complex conjugate of ¥(/)
Let,
XD = Gn (13)
OT agro” (aay Cree art Pe
Substituting equations (13), (14) in (11), we pet
Taf
Ynray
ae
ee aan my
rn
sree
aa
BAU URE Chay
Lee Lea Es (eat
DSRS ene (1) can be writ
ene (Perr xf
* From equation (10))
Sele AU aE CeAyP
Pal Ole io; ,
Bes: = firoiy- fey LAU Drel GATE
ace Tete eaten
ay
[+ From equation (14))
warnne: Xerox/Photocopying of tis books @ CRIMINAL acl. Anyone found guilty is LIABLE to face-LF at a.
Scanned with CamScanner oS
the matched input
the auto
cern
gnal
Proof
Soe AU Aen eae
Eire tonsone tre ne ee
oat
eee A)
and energ
cesta
GI
SURaeP day
Rare
Ome ners tr eras
oo
The matched filter output signal to no
Sure
CO “
OMT eSTnyE pee RO cea
spectrum of the White noise at the input of the filter.
Re Ue Se eta et Petre Rute rt te
ieee ne
5
(SNR) a XP) af
And the ene
rey of the signal,
B= [s@ar
From the Parseval’s theorem,
fs?@ae
J S(f)) df
Then,
Jewa
E 2E
(NR), ==> We
2 2
Hence proved.
The output of the matched filter is same as the energy
of the signal at the filter input.
Proof
As,
BD) = [Hap 9.) OM of
Since, H._(f)= S(-f) €?*” (for a matched filter)
Substituting the above value in equation (2),
‘Then,
Sif) = scr sy PT et" df
aS
s= fISP a <
87
What latched filter? How it differs from an
optimum filter. Derive an expression forimpulse
response ofthe matched filter.
(Model Paper-4,Q6(a)| March-21, et.2,a4(0))
(or)
Explain why the matched filter is called as an
‘optimum filter. Why the name matched filter?
OctNov.18, Set2, a4
(or)
Derive the transfer function and impulsi
response of the matched filter.
Ans:
An optimum filter is called the matched filter when the
input noise is white Gaussian noise.
The psd of the white Gaussian noise is given by,
eg=2 ol)
We know that, the transfer function of'an optimum filter
is given by
P*
GO
‘Substituting equation (1) in (2) the transfer function of
the matched filter is obtained as,
pe
a
H) =
pat 2)
HO)
a) 2 paipem 3)
ID) = Pe
Applying inverse Fourier transform to equation (3), the
impulse response of the matched filter is obtained as,
j
2k
HO= | Perera (4)
A physically realisable filter will have real impulse
response not complex. Hence, h(¢) = h(t). Then equation (4)
becomes as,
M= = [ Pper-nap
~()
5,(0 equation (5) can be written as,
2k
Wo= P=)
Since, p(t) = s,(A)
nO = si [s(7-9)-s(T-9)
Impulse response of the matched filter is,
2k
MO = InP) -s,O= 9)
The above expression represents the imy
the optimum filter which is a time ve y
of input ie, itis matched to the input sig
matched filter. Tae
Scanned with CamScanner Case (ii): 5,9 =— 8.0
with s,(0)
‘Then equation (4) becomes as,
ee
eS
¥y of error of optimum filter i What is correlator? Expia — F
a. ‘Plain the optimum filter reception using correlator.
(ede Paper 3 Oa) | March, St, OM)
(on)
Draw and explain the
coherent system of signal reception. (Apri-t6, Set, Ota) | NovsDee2, Sat2, Ma)
. (or)
With neat diagram, explai
plain correlation receiver. Compare correlator and optimum filter,
Ans
A practical realization of the optimum re he coh
ceiver is the coherent detector.
A coherent system for signal recept
ception or a correlator‘receiver is'as shown in figure
Dn Oe Oe AOS Cc a
rennin
Omer)
t eae
ge Lee et ee
omparison of Correlator Output and Matched Filter Output
ont iched filter as shown in figure (2).
rea od
Rett
‘ALLIN-ONE JOURNAL FOR ENGINEERING STUDENTS |
‘Scanned with CamSeanner (Data Transmission)
ie ee
@psk DAL eR ee ea ee ea ara Sasa
Se
Peano hari
cease
CTR Una et ee
eee
Peres cn pet
Sa eety
Signal to noise ratio, S/N is less
Biers acest
islow. . 8
Loe ako ane
TSE Secs ren esa
inne
Cane enen
Absccn
. Derive the pro
The signalling waveforms for BFSK modulation
interval 0.<1 1d 0 elsewhere
The coherent BFSK system is characterized by at
two mess
1¢ points can be determined as follo
The coefficient s, is given a:
5, | s(oo(Oat
Where
4(0 = [Fr eosanft
Where, s(¢) is the general form of siz
bcosdnft , OSS
0 elsewhere
s{)=
T,
ne are s\(f) cos 2mft and s,(t)
erence ere
Pearce imei
Due to the absence of additional circuitry, circuit
is less.
non ae Rona
Ry ieee ca eco eee
is high
Bandwidth requirement of non-coherent system i
eect
Ree oO Cnr care
applications like slow speed data transmission and
‘medium speed data transmission.
Lene ee
eee)
Ce ee emer)
eee
® cos 2nfst over an
Vt Vn
» dimensional signal space, which consists of two message points. Tht
ad
spalling waveform and is defined 35,
+> gpecrnuM ALLAN-ONE JOURNAL FOR ENGINEERING STUDENTS anes 4]
Scanned with CamScanner By substi
sang af) fom equation ()
‘equation 2, i equation (D>
$270
‘Similarty, the elements of signal
f= Oands,~ VF,
ene pesos sve
lal ste (]
In the presence of noise, the received BFSK signal (0),
‘canbe given as,
af) = 5,(0) + n(0,if's,(0)is transmitted
=5,() + n{0), ifs,(0 is transmitted
Where, n() is the white Gaussian noise with mean = 0
y,
cand variance =
e
[
VB}
1 BB coon
Th i,
and o(?) from
ivf
ang Xare 00nd YE respectively
x0) = 0 and E10) = Ee
cd on the transmitted symbol, the mean ¥
ind as follows.
he conditional mean value of
nsmitted is givenby,
Bases
random variable-L is fou
‘The expression fort
ble L when symbol 1 is tra
AL) = 21x 11-20)
~
Vo
conditional mean value of random vagy
Similarly, the
L when symbol 0 is transmitted is,
EUW/0) = 1X /01— FLX/91
-0- JE, =-VE»
since X; and X, are independent random variables
variance of Ls given as the sum of variance of random vata
and the variance of random variable x(&)
ive., VarfL] = VarlX,) + VarlxX)
No N
‘Assuming the transmitted symbol is *
» the expresin
for the conditional probability density function, is given’,
i rey
fun @/M) = exp( - C54 EB)
V2nN, ?\ No
The detector makes the decision in favour of symbal
for x, exp| gy | IFAR>>N,,
Ans? ee
Non-coherent ASK: Non-coherent ASK scheme doesn’t require any phase-coherent loca
@fASK signals include rectification and lowpass filtering atthe tec
forthe ASK signals. Due to the nonlinear operation of the receiver,
cillator signal. Non-coherent operation
ck diagram of
ver, Figure shows the
calculation of error probability for this receiver is difficult.
P70) x =
Figure: Noncoherent ASK
The receiver input is given by
[Acosw,t+n(Q) when by =1
Wi = @
) 10 when
Where
n{d) —Noise atthe receiver input and which is zr0 mean, white Gaussian noise
Fol bandpass filter, which's centered a, and having a bandwidth of 7. The ouput ofthe filters given By
ORSON)
Re ae Ue
Pea
Cer rr itu aaa
Scanned with CamScanner Since, OC)
be written as
© at the receiver input must
Welop detector above the noise Substititing dt, andle ‘tte receiver input, threshold sting,
sfthe cori strength andthe optimum tinesold sti 7
‘Then, the probability of error is given PY>
Then,
| 7, |
| rf a |
1
But, fora PSK system
5, - and £
Wher,
§, = Average signal power
£,,~ Signal enery per bt
Then, the probabil
be written as, mee ore
roo
Peon)
r
pIGITAL COMM ——
Seer
Ens
ee OCs
en Le ae
Mn tu aa
Meer rt Ara acal
perenne
ape Ua
May
Se
Penta os
OU eee
S( =A costo," 0S 1ST
Bit duration, 7, = 0.2 msec
mere ts scoala
eee ena
Sa eel OS
Average erormic=? >
eer ac
error is given by,
Varad
nd P,, fi has to be determined. $9}
Jl2-soPae
$44? cos? wre
WARNING: x
3: Xerox Photoc
opying ofthis book is a CRIMINAL act. Anyone found gui
tyis LIABLE to face Let
GAL proceedings.
Scanned with CamScanner (i) Non-Coherent FSK
‘The probability of error for a non-coherent FSK is
siven by,
“)
= 0.0088 mW
Thus, power requirement of non-coherent FSK is
0.0088 mw
SoU one eare te eel
with usable bandwidth of 2400 Hz using the FSI
ee eer
2025 and 2225 Hz, and the data rate is 300 bits/
Ue Ce Cnet eae ee
the output of the channel is 6 dB. Calculate P
Sea euneaett aee ter et
Cae
ama Ita ener
US a ees ed
eS tee a aaa
2025 Hz and
Do ae ee
CaS ae
arth)
As, f ce ee DLA
A ea
The following conditions must be
ena
PT era ra
.T,?* Vs 0,20, and 20,7
ie,, 0,1, = 2m x 2125 x
44,5 >:
300 - !
0, >> w, "L>>f)
1
20, 7, = 2m x 2 * 100 = 3m
ty ame AS
2
IGiTAL COMMUNICATIONS (iNTy,
aoe rarcoherent ASK scheme,
#
nal power, S= A
‘Average signal po. 2
=2xNx
Noise power, N,=2.*31 x 2400 = 249
Average signal power
hen, AE
A12
2400” 48009
a = 19200
7
A& _ 19200 _
Then, 300
Aes
[6.248]
Then,
2 (6.248)
= 2.129 x 19-10
[2 P= 213x097
For non-coherent FSK scheme,
The probability of error, P-is given by,
2 ¢xpl-8]
1.677 x 194
© WARMING: xeroPhotocopying of this.book ie » oa
Scanned with CamScanner pee clar
nga
Ca
PC eee ST
ee ee
PR Lest re
pene a ees
error probability when
iad
eee ee ee
Sats ae
eee er oe)
ford<1 <7.
AO eee oe
Do aces iret eat ee
RE cee. (Ogee)
Pe O ier (Oren
