/

OPTI CAL
;

i'

f.
a

u
s.G.]GRAcg :SHoBA
..: .

Department of ECE
Velam mal Engineerlntg Cottege

t
j:
.l

Ngvember / Decembe r - 2OO7 1-39

I
'1 .-''28
t
!*
November
/ ?d.ernber . 2008 ' 1-27
", 1-27
Novernber/ December - 2009 1-30
1-43
November/ December
\
- 2010 1-36
t
. -,33
November/ December - 20,l I l

1 - 3g'.
May / June-2012 1-59
November / Decernber - 20 12 Refer Supplimentd Book
.,
 - .-ia.iq-"f *i._.
!G..

s
. r.l{

R 3324
B.E./B.Tech. DEGREE EXAMINATION, NOVEMBER/DECEMBER zOOt.
Seventh Semester
(Regulation 2004)
Eloctronics and Cqmmuhication Englneering
\" :" '.'
: .

(Commen to B.E. (Part-Timel Slxth Semester Regulation 2005)

Tlme: Three hourc Maximum: 100 marks

,. PARTA{0x2=20merks)
-' : :,
':

Ans: For monochromatic light fields"of radiation frequency c0, a mode travelling in
the positive Z direc*ion has a time and Z deirOndence gr* bV
4,,'pa
SZ component of th9 wave propagation constan t K = 2rrl1, and is the main

2. -.A silice opticel fiber hes I cure refractive index of 1.5 and a cladding
refractive iridex of.l.4?. De&irminc the acceptance angle ln air for '

the fiber. t

Core refractive index 1.50 (n,)

Cladding rcfractive i ndex 1.47 (zr)

The critical arigle, 0;=.sin-l

Ir)

I
 --:-'w.+. Ef,
&
? I)esree Exam inaEon g
Acceptance'since in air $a = sin'r 'N.A

-
3. What do yqu-.t mean by potarization mode dlspercion?
-8.E./B.Tbgh.
gns: 'Polarization mode dispersion'means the dispersion due to the removal of
degeneregy between the trrororthogonally polarised modes irsinglc'pqde fibers
which causes different phase coqtants.fur thbse wo modes.
i
4. 'Dtstt*gUlsh dlsperston shlfted and dlspersion fuIttemil'frers.
Ans: (f) Bbpersion Shllted Fiben
Normally at 1.3 pM, thcre is minimim dispersion and finite loss. But at 1.55
pM, the loss and dispersion arc very minimum for siQ fibers, At 1.55 pM the
ma&ria!dfisperslon is *vs afld wavOguide dis-persioi is negative.'Whcn these ar.e
equal, there:is no dispersion. So the-operpting wavelength is now shifted to t.55
pM insteadof 1.3 pM.
.; -
k'. S. Wnet do,you meill bi lieteroiunction? Mentfon its advantages.
'A heterojunction is an interface between two adjoi[ing single crystal
Ans:
i'' semiconductors with differerrt bandgap energies. Devices.that are fabricated with
' heterolunetions are ,said torhave hetJ*r*orJ." ''. '
.,
'
' 6. List the differeat typesof mechanicrl misslignmenqi $at can occur
between two Joined fibert.

0i) Longitudinal (end separation)

7. Whet lne:the deslred features of r phofodetector?

(3) Low dark cunent.

, Ans: er***lr*t in t r Uigittl trirnsmisiion i3 ttre minimumoptical pulse enqsy

*E*- required to maintain a given bit enpr mte (BER) which any prac*ical reeeiver
I must satisfr. I.n the analog Uansmission, qtiannrm limit is fte minimumopti,gelpOwei
.:
r
I
9. lYhe! are the advrntrgee of WDM? ,, :

k
'I Ansl Many dilTerent wavelength can be sent along a fiber simultaneously in the
1300-1600 nm spectnrm.
L

l
I

'
f
r;
l rO. . Disringuish fundamen,"f "n
B
i-g

! 'lns The pulses that do not change in shape ire called fumlannental soliton and that

' ' ,,
|undersoperiodicshapechangesarecalIedhigher.ordersoliton.
'.ra11ryas*16=ggnn"rksy.; '

I ,r.OXt)
,,
I prrra! withr!'f-ert'dlagran rhe etemente ot an opttcdt tiber
!' I i ' ' (lo)
tf'ns: Besic Opttcel Filter Communication and vsrbus metricrl of Ftbcr
I Optical
II Transmkslon Link.: ' r

. Figpre'(l) shows the basic componGnts in the optical'fiber communication

t
I

The input elecnicalsignal.rnodulates the iqtensity of light fiom

the optical '

sotirce, Thc optical carier cui be modulated intemally orixrernally

using an
elecEooptic modulator (9r) accoustooptic modulator. tto*. auvr rtt r.
modrrlatqr-s are wilelf
3sed
q:69 ""irr
externaf modutqtors whichtrodulatc the hghl
: :.
bycgBng Jr#r"r r.r**,
refractive'index'tl,ough the given input
$e
The tansmission channel is optical fiber unit. At the receiver .
end, thc carrier
, tv.demodulator.,rtrus &eloriffi ri;ilil;."r,;r*ii]r'*.,ur.a from the
 --B: .: -.w,'*
1 , FtB
'i

4. , .j . R.
^ls

-
B..E.lB.TechiDerrEsEelrirradon.j,
t l;(a)(iilA mottimode step index liber with a core dieineter of 80 prn and I
reletive index difference of 1.5% ftr operetlng et a wavelength of
O.8SgLm: tithe core refractive lndex is 1.d8, estimatethe,iiorinalized
frequency for thg tiber and the number ofguided modes. (q)
Ans:
- Coreaiame€r
-------------.-. d=80 um
"-r--- ''
i'
:

i.R'elativeipdiced|ffeq,en9GA=l'5ou6=0.0l5:

Normalised frequency V = n NA (or) n, JzL

* "*
.ffi,,-,75.8
,i

. tr, (ZS.gf =2873

No. of euided mooes =
-=-
ff. O) (i) Discuss the evolutlon of liber opth c:ommunicatlon system. (6)
'iAns: In generation I, mounting Gals'faced LED's ard laser diodes having
g corission wavelengths 0.E lrm were used. From 1974to1978, graded index mglti.
-ro{e fibers wcre used. From't9?E onwar&, only sinlle mode Jiberq are used for
;

During the socond ge,neration, the wavelengflr is strifted to 1.3 pM to oversome

and dispersion. Further ImGaAsP heterojunction laser diodes are used as
I sources. In the third generation lhe operating wavelength is firrther shifted
l.55.pm and the dispersion shiftcd fibers are nscd. Further single mode direct
ion is adqpted. In thefourfi generation erbium'dopqd optical amplifiersrare

,i
In ths pro.posed next ggneration solitgn basod loss lcssand dispcrsionless
l the data rate may
increased d 100 Tb/s.
i

Wavdength Bii Rate f,ass F.xiited

of Optical (Mb/S) (dBrkm) ttpto

0.9 4.5, 10 I 1980

1.3 1.7 x,, 102 ,50 <t r 987
1.55 1.0 x l0{ 7A <: 0.2 1990
1.55 1.0 x 105 100 < 0.02 2000
 *rrrolr.
*
.F

dotic al Comm ttnication

ir5+-r*ri--#-

It{bxii)Sketih' and elplain the electric nete,distribution of,the lower

'..tr$ergutd9dqodesinsymmetricalslabwaveguide'
Ansl Flanar,**vegl.lide is a qim.Ple-St opticals waveguicle, it cursiqts of a siab of
A"f".Oi" with redctive index n, and rund*i.h.d between two regions of refractive

Figure shows that aplane rnonochromatic wave propagating in the direction

Optical Wavclength within the gqidg is reduced by the amount n, and vacuum
p*portion constant (*) is indreased by refractive index n,.
ois angleberween the wave propagation vector. The planl lavelan be Ttlluld r

into two comptlnents orn is popagatlng in 8re ldircction drd thecther one is in the I
Z-direction. The'cornponents of plrqse propagation constan! in the Z andX,dilections
are given by,

Pr- nrk sin 0

Warm Vechr'0, K

Fx=o,KcosCI

Fig. a
X Dlr€ction

I I .l ! cr*,q
I
I
':l
r I
l--
I
I
I
lr
t | 'E'F|BE

Fig.b

i
 -r-etr:,*%f,*
I
*
.F

6 B.E./B.Tbch. Dcqnie Examinatibn .

Ttre X-direction component is reflected at the interface between lower and
higher refractivp index. The phase after two reflections at the interface points P
andQ is equal to 2 mz radians. where 'm'. is an integer and known as ottleT-ol
the mbde.
The constructive inteffertnce occurs anda standingwavepattern is obtained
inX-directionTheinterfererrcetennsttrelowestorder*oarrt
.I rdingwavepattern
fto interference forms the lowest order mode
is obtained in X-direction (Fig.'Q.
t standin€ urave whercm =0. Though electic field is marimqr.r atthocontre, it
penemtes some distance into the cladding. El'ectric field distribution in the X.
diregtio.n &es not change as the wavc pppagates in the X-direction. The sinusoidaly
. varying electic field in the Zidinection-is showa infigiire 6. The stable ef*t i.
, field in the X-direction withonly a.periodic Z-dependence is known as mode. As
snowq i* ri$tr e d specific mode:is obtained onty when the angle betrreen the
1 Propagation vector and the core-cladding interface.have a particular wave. These
modes have a periodic z-dependerce of the formerp{-jPrz). If exp (7'orl) is the
' time dependence of the monochromatic electromagnetic light field with angular
. frequency ".c,)" the n exp (cor
-BZ) desgribe the mode propagating in the Z-
dirrction whbre BZ is the propagation constant is irivariant for the ritxh',h"o.e
W=b-
Uten tight is conside,red as an electomagnetic wave it cortsists ofperiodicqlly
rarying electic.field E and the magnetic field .Ir which are at right angles to
* cach other. Flg.c
Figure c shows the appearance of niodal fields in the planar dielectic slab
f:rrrveguidb. Patterns of several lower order tansverse electic field mode (TE)
i'rc shown in figtue; Hgp the electic field is perpendiculqr 6r tansvefse- to the
,"
fioction ofpropagation'and henceE = 0 but conepondi4g comporcnt ofmagrretic

fu* F in the direction'of prolagation is not equal tozeno. Here the modes are
'
t*t to te fansvenp. The order of mode is equal to nrimber of fi eld zero's across
gorac. The order also depends on the angle g '. steeper the angle,higher the
;1!:
-grdqofrrode. The fielcs vary harmonically i.e., sinusoidally in the guiding regron
ffrc&active index n, and it doca;s exponantially outside the region. Fields ane
iltiUteA toumrds the edges of the guide-for higher order modes and for lorner
' .tr.odes the fields are Jonnnea-,i*,lr;;b.r;r. Tha fi;ifi;ffi;;
t fiber acceptar-rce angle is refiacted outsid€ and rappeil into the ctadding anp
be result in cladding modes. firese rhodes also known as radiation modes and
'*lleboasolutinnofmarwell'soquation. Asthccoreandclad(ingmodespropagate
.Fqg te t*r the r4ode cotpling takcs place . uode couplini focc,ree Uetwlen
ry-
 _.sr*;*t,*%f,F

*
.F

Opiical: Cornrn unication

higher order modbs and thc radiation _modcs, Diffusion of power back and forth
. between arc and cladding modes occurs and results in power loss. ln practice the ,.

cladding rnodes are supprcssgd bya lossy coatirig. ,

A mode remains guided along the cored as long as p the.phase propagation

constant'satisfres the following condition.'
.. i

"' n'k<P<n'k

Once B< n,lc then power leaks out of the core as leakymode into the cladding
negon These tqgly modes ..c.qn c5y srgpjficant arrouni ofporyer in short frbe*:
Leaky niodgs ar€ $aqtial{y confined tb the core gegion and attenuated by
continubusly radiating their power out of the core as they propagate along the
fi-b6 fiis.power radlation ii due to tunnei .flba!. The cut oIf cilildftlon betureen
leaki.mode arrd guided mode is S<n *. i , ,. ,l

TE I TE 2
'lr I
.t I

ffiding nr I
T
I

Fig. c
F '': ---------"q [t'
1. . .t
8,, ',:. .1. .: I n.H1g.t"gtip***.e*"*ioa.iorrf
t1.(bxiii)Draw-the structures of step index rnd graded index flbers with
their typical dtmensions.l
Ans: structur.e of step and graded index fiber (f) with their typical dimensions.

fll Step'index fiber multi.riode (2) Graded index ffiutti.fiode

Figs.
'

(2) Step index multi:mode-fiber :

(3) Graded index rnulti-mode fiber.

l

e)(i) When the mean optical power launched into en 8 km length of

fiber is 120 lyfW, the mean optical power at the liber output is 3

(1) O?erall sigryaf attenuation in dB/km and

(2) The overatl signal attenuation for a f0 km optical tink using
the same Iiber with splices at I krn intervals, each giving an
attenuqtion of I dB. (6)
 .-r<B-.*i-
8.
,.9
.8,

tical Communication gr
'Ans: P '-""'"
l20xl0 6

Signal attenuation = l0 log il=to los

- : 9ou, ='16,0 db ' :

16
Attenuation oer km
=
i = 2 dB/km
'=
Atteiruationin l0krnfiberlength, Attentiontttm x 10 bil
I .=2
x 10

- 20 db.
Totallythere are 9 splices in the l0 frz link each with attenuation I db. Toal .

db - ,
,

.'.Overall signal attenuation = 20 db + 9 :

= 29 db.
.:
rchanisms that
.ort"ibotu to attenuation in opticel fibers. (10)
Ans: The,aftenuation or'tansmission loss is ong of the important pro-perties of
o,ptical fibers becalrse it detennines the maximum tansmission distance prior to the

Transmissionlosses.offibers were ieducedtelow that of inetallic conductor

i.e., less than 5 dB/lorr.The signal attenuation within optical fiber is expressed in
logarthemic unit ofdecibel The decibel is used forcomparing two power le1ls,

where, w=; t
L;G|J
:.

LK"') z ",-LpQ)j
[4Pl= 4.343 ophma
"|-9]=19 '
In cptical comnnrnication, the dttenuation is usually exprepsed in decibels p9r
,*ritlength. ': I :
 -,"=qg.*{%r,
''
s
"l+

Pi (decimal/length)
a dBL -'10 log,o
i
Po

Standard Fiber
E
--
oo
.t
g
.g
oEt

=
e
o)

800 !000 1200

Mechanisms are influencedby the

following factors;" .
.
runnca technique,
(1) Material t:*position, (2) Preparation !
and Purification

(3) Waveguide A .
(1) Absorption, (2) S"uttoins, (3) a"::,i:d Microlfendilt'.otttu'
(s) Losses due to teaky
(4) Material coupling radiation losses, Sodes:
silical glass fibers:
absorption losses in the
A A
E E
-- --
a
ar?

4 0
E
4l
-,
=,
o
\J
a 3
o 3 rfi
o
J
o
J 2
2 E
o
=,
() o
(J 1
I
=
T'
E 0
€= 0E
2m04000sosffis
Time After irrdiation
Dose (rd (Sil

Fig. (b)
 -l-';aa-t,*ilWG..

s
'.F

The imperfections may be missing rnolecules high density clusters of atom

groups or oxygen defectq in the glass structure compared with intrinsic and extrinsic
aUsorptiop this absorption are negligible . tfthe fiber is exposed to ionizing ?diation
which may occu, in a n rclear reactoi enr;ironment, in medical radiation the rapibs,
in spacimissions that pass through*he earth's van alien beltp, or il accelerytoJ
instnrnentation. Dueto the radiations internal stnrcture ofthematerial gets changed.
The total dogs received by the material is exprqssed in units ofrad (Si) which is
a

measure ofradiation absorbed inbulk silicon'

' Thi'unitisaenneltti*i$= j' I I

' l.,. ' .

=o.orr* gflee"
,fire damage gffects depend on the energy gfjodzing pqlicles br raVs sulh
ry
elecfions, neUtong, or gacrma rays, radiation -n* TO the fluence. Attenuation is
: : increased due to ionizing radiations causes attenuation ccntres that absotb optical''

e|lergY. , , : , ., .. :
, .,,,, level, larger
.

fr€ attenuat-ion due to radialioin i.s an induccti ioss. Highertheiadiation

,
the afe-nUation and the attenuation cenfes wilt relax with tirne as shown in figure
.t'

IntrinslcrAbsorpdon: '' ,

G) Inuinsic'absorption loqs due to ultaviolet absolption tail, :

(ii) Intrinsic absorption loss due to infrared absorption tail.

The peaks are brord and ail off inlo the vygus nart olthe :p""fi'fT: inf-re-nt f[
interaction between the vibrating bor-rd the
, asbsorption is because of ttre -and
;;;sretic field of the optical signal resulti in a qansfer of enerly froin the
Ecld to the bond, there b1SivinqrisLto absorption. This absorption is st0ng becausu
.

of enormous bonds"are present in fiber.

- "rhc
uredge shaped specfial-ions characteristic for oIR is shown below.

f".
 lnfrared absqrption takes place because photons of light eneigy are absorped
by the atoms within the glass moleeules and converted to random nrcchanical
vibration'TIiisIRabsorytionatsoexhibitsamainspectral'peak..
An emprical expression for the infrared absorption" for Georsiorglass is
gvenby,

fl/ft = 7.81 x .10r' e(:48.48)

, llavelength (rrm)

I
-
E
..-
tfr
ilEl
-
,
/, I
C"
C'i,
o
-l
\

0.01
2'.5 - 1.5 1.O O.5

FIg.
Totat lqss spectrum fora tlpical optical fiber:

: 154'2x
--uv
o....='46.6x +60 -lo-2rfjg)
- -[ ],
)
Oo ticat Comnn urtlgedgn

'

Ftg.
'lt

The introduction of impurities such as germanium dioxide to

modis the-
o{ry.absorption tail because of
. refractive causessome in'crease in the ry1q1ilude
takes place'
an upward soft of the wavelengtti at wUi* tne peak absorption
Extrinsic AbsorP0on:
e*rinqicabEorptionis&retoqqngitlonmemlir"pruities;Tlstlneofabsorption
is dominaniespgcially.iriAe frUers prepledby conventionat ryftingtecliniques'
table ythft:
The commonmetallie impurities founa inglasses are shovrn inthe
absorption losses caused by one part in 10e. some irirpurities
tike chromium.and
.* cause attenuation in excess of I dB lon-r. The presence of impurities
"oooo
*"i* a"prq,s andb. thertransitiqnmetal gi6sriseto additio,nal absorption inthe
longer
,iSible ani'near infiired rurps. Also the imprrities shift fte qabsorption to
wavelength.

050 0.80
&(pm)
 '..8
s.
14,, , ::
' The figuie illustrates various absorption bands due to Cd-, Cuz", Fd*;"and
0H'orhydroxyl ions.
Peak One Part in lOe (dB Km;-r

Cr3* 62s 1.6

]

c2+ 685 0.1

Cu2* 8s0 l.l
Fe& I100 0.69
Fe3* 400 0. 15
'
Ni2n 6s0 0.1
Mn3* 4oo 0.2
v4* . 725 2.7
The apsorption ppeqt*ry for the hydroxl (OH) group is silical.

E
;
F

-.Eo
c
o
:F
g
,E
.s

overtones appearsalmost&ornonically at 1.38, 1.095 and 0.7j pm due to the

fundamental vibration.
12.(bxi) Discuss in detlil intramodal dispercion with relevant expressions
erd diegrams. (10)
Ans: Intramodal Dispersbn: Intrarnodal dispersion arises due to the dependence
of group velocity on the wavelength. Further it increases with the increase of spectral
wid$r of the optical souroe. For exarnple in the case of LED, it has large spucral
width about40 nm since ifemits wa{tlengths frrom 830-870 nm. withpeak emission
wavelength is it ASO nm. In the case oJlaser diode which has very nanlow spectral
widtltr:about I or 2.ryn only. rhus the intramodal dlspersion can be reduced in an
optical fiber using siirgle rnode laser as an.optical source. Intramodel dftipersion
 -.F.wrr:,%f,n"
'.F
,. g.

Qe.ti+"tCpqfo,rrl"eli1, ;,. , ,, , , ,. , ,, , ., . ,-:. 1f'i,

*i* Au" t1t thi dispersive prgperties of the optical fiber material 't'l
and the gqidance
'

(n)rMaterial,pispersion (g1j Chromatic Dispe5sion: This:dispenior arises due

to the vanttion of.the'refraitive index of the core material with the wavelength or
frequensy.of lightj It is directly proportiongl to the frequenqy bandwidth of the
Uansn + p.lb* *'^.Ldrbi=*eC"tde"sir, wlgrt dl, t d! * O'

da
: 't -- dp'
.But Fr :

The prlse broadening due to material dispersion is obtained through the goup

I /v" t( -^aJ
=;=t= ^da\
... Group delay t" ,[n
.,.
t( - dn\
)\rer a l*gtl .L'- .,: ,^=Z[.* - ^aJ

r.m.s. pulse broqilening due'tri material dispersion =.o*= o 1, *

Her€,.,^
o, = rms sPectal width

' fL(
= mean wavelength oflight sotrrcg. :
X.
*t.

o^=",
&t.ql)]="**l#l
au,, A
1;[.#-rari-rt)
L.

r: Il-dr*
Material dispersionparamter, M- +=:lHI
=Llq'",|

:
It, d'nrl
Tr

"f I
-ttY d7"I
'M' has the unit Ps nm't lcfift"
 .r,l[
s.
.I
tlon ' ;*

Wbvelen$h (Pm) '

Fig. UUavelength (pm)

puz' the material
-*1 Fisure shows the variation ofdispersion with wavelength upto 1'3
is positive, At t.3 pm, it is zero, Beyon! 1.3 Mm, the material
dispersion
fu.p#or
J.
$Fnegauvc
o)w""".GuideDispersion:Thisdispersionarisesduetothefinitefrequency
ffi;;&r, *d tu" a6"na"nce of the mode group velocity with thc frequency of
ighr Higher the frequency bandwidth of the transmitted pulse, higher will be'thb
li**Ie dispersiotr. fh!amount of dispersion depends on the fitrcr dc-sign like
ldiur,,rina" the propagation constant 'p' ib a function gf all,. In the case of
case of
insle mode fibers, waveguide dispersion arises when d21t dX'+ 0. In the
.themodes
fibers, most of propagate far fromthe cutoff value.
r - -a: ^--
ry of waveguide Eispersioni Let'D'be the normalizbdpropagalon constant
a mode and is defined as

.8,,8
L:-ll:" *- - fl,
.D=.7-=-
,i
T
,o
9= ; ln,* (n'- n')bl

n'1ul+9 6' -''#

#=t=:'ln' + -
(n'

\
*
Since D is a drnction, of V number of the filter and Znumber is the furtction
F of,co, I

dBl:
L---
'aQv. da vs

Thup the group.vefocity depends on 6) €Yen in the absence of material

 I .-"**,"il%E,
I

:
\ g.
I ') ar'.
,Ootical Communicati,on .f
. .. djspersion.Thisdispersionmgchanismiscalledwavbguidedispersion. , ''''
.'
' 'r ':
.;
1,' :
y*{ o,[*i,i=?r,[rm
Aq
. \.q
tt
v
I dY a r-'Y'-'
=-O
I
dOl a Y--'
-=-rln.'-n: '

' ' tl, ..(rr, -A\ d(bY) 'l

ccdY
., -....
.1: .'_-L L-1,.M(bn1
.'. t=-=-rlril+# i

For Slnqle Mode f ibers: The rms pulse spread over a spectral spread o*,
' dr_- . L .dzfilildY
o*=o.'t'=ore4roffi,
-rm U
^q
\:avt -r
n
L .lra'vu1 : ('.av
I ' | '.':-=-- Y).'-
,.
=-6. .-.2. Al =#
,^c1.. L dI/, I I

, \ d?l, L)
orq

-h-. D-A is th'ura.,€Eiide dispersion parameter and is eqnal

^j-'#')
Takin g Y = l.Zwheri;p;* (I.) reaches rnaximum value *

(i.e.)

' For [ = 0.01, tre = 1.5 and D*B(I) = 0.2

\
9** =- 0'003 o^
:

I.
 rl?i"_,.3-'*1Wf,iF.,

*
..F

;, o^ [r, .d'n,l_q.o2o^
..:-:=:-l ,\. ._? l___ _.
,

. L grl',-
=
d7" ) c7\
.

c-,> o-.. This is tnre for,shorter wivelength. But fbr longer liavelengths
rIwS:'
6:mwg
ac 6.

ft us take a singlg mode fused silicon are fiberl The variatio:rs of materi.a

c
I
.8
}A
(l,
o.
.g
()

If is observed that the total dispersion qan be reducedto zero at a particult

wavelength due to the cancellation of the mgterial dispersioh anri ncgativ
waveguile dispeniion, Thus the totatrdispersion can be reduced io a minimur
value in the range of wgvelength 1 .3 pm- l .7 pm thrgugh the proper fiber desigi
12.(bX1) Wrlte a urlef noti on design optimlzation of Cngle mode ttbers. (
Am: Since ti.he single mode fibers are widely used in the telecomt4unicatic
plications and optical,networiks, the dispersion produced in a single rnode fiber
pactically very importarit since the dispersion reduces the maximum hit rate t
. tumission capacity,as n .11'6s thg $andr ridth of the fiber. Zero dispersion in sing
6gde fibers can be achieved at the operating wavelengths by the suitable deisgn r
re&active indexprofile andmakingtheaddition ofmaterial dispersion and waveguit'

(r) Reilirrctve lndex protilesr (i) 1300 nm oprimized fibers.(i!) Dispersion shifte
n'o.rr.(iii) pispersion flattened flbers. ' . ,
 -
==;;p.'"t- G,
.,9
B.

Optical Gommunication 19 '

Core
,2pm Az = A12o/o

' Thus its refractive index is in the form,of a depressed nmnned sunoundin!
with high refractive index regions. This refractive index profile design is to optimize
'tre waveguide dispersion so as to get, So as to get zero dispersion at 1.3 pm.

= 3.1 lrm

a = 2.2pm ?z= 4.0pm

prn
I ", = 5's
Fig. (a) Steeper step index Fig. (b} Triangular with annularing'
The refractive irtdex, variation is in thq formof a dip in thg regron between.
wo successive ma:rima of refractive index of core
ff) Dispersion flattened fiberfi

1 oi",d

l/
l=
A: = 0.457o

.l.ai=3.6pm - 8r =.3.4pm
&-----1s
l:__l + = 4.tpm a-

, ,.
. 14',a.' .:
:

F(1. (a) Double clad or W profile Fig. (b) Quadrupole clad profile
-
 -.-"+s'"i"-
l' :.
:- j. lf,G,

s
1 '.F
I

B.E.E;Tech, Degree Examinatiot?

Signal degradatioq inoptical fibers:
. Here the refractive index profile is slightlymodified from dispersion shiftect
frbers.togetzerodisperrsionforawide.range: . ' , ' , "
1300-nm
mised ' Disper
Flbtte
'20
E /
-I10
E /
e-
e0,
o
e
.B,o
a

1.3 1.4 1.5 '1.6

13.(a)(i) Draw and explain the LIED structures based Double Hetert

Ansi- Lrser X'undamcntals and Laser Mater'rals: Laser is an acronym for l"igti
Amplification by,stimulated emission of radiation. To produce optical arpplificatior
and an opical resonator to provide the nocessafy opticat feedback Laser Actior
Eeans the amplification of light by stimulated emission of radiation.

'

E,

(a) (c) (d)

Figs.
(e) Absorption: By this absorption process, an atg11t in levelE* absorbs photor
of frequency (Er- Er'/n and goes to upper energy level Er;
' : a..
"

o) Spontaneous Bmission! An atom in the energylevel E', i,iln make transitior

i!

to lower energy level E, spontaneously ernitting a photon whose energ'

equal to 1nr- Er). . ,
Ctrrecteristics: (i)It produces polychrsrratic radiation.'
 .F
a R.
s".
)ptical Communication

".
, (ii) It's intgnsity is always small, (iii) It is not coherent, (iv) tt has'lowrange

O 3ti*otrt a'Emisslon: en atom is an exciteO,level or higli energy level li, c,p

.l\

ralS a nansition to lower energy level E, by an external photon of frequency '

E, - E,Yn. The stimulating or inducing photon.

-:haracteristlcs:
(i) It produces monochromatic radiatioq (ii) It's intqnsity is very.high' (iii)

Nl
Er'i - ' '
TT,

a
" 'E -8, = + =2.86 x 10'tV
aa
h
KT at 300 /( = 4,14 x

If the ernitted wavelength is 694 llm.

.'.:

J
KT at 300 ft - 4.f 4 x '1,0- 2r"f. .t

lb.
N,
 -.=q[*o.,*iiiry[E..

s
''F

Constructlon:
ln GaA,s diode laser, a layer of GaA,s is sandwichod between two layers of
which have a wider energy gap and a lower refractive index that Gals
fraAl,a,s
$Carls subsuate - similarly inInP diode laser, ala:reroflnGaAsP is sandwiched
letrreen turo layers of InP which.have a widc energy gap and a lower rcfraotive
hdex than InP on InP substrate.

D The basic principle of working of hetero junction laser diode is sarre fuith '
, respecttohomojunctionlaserdiode
i.
qF The GaAs diode has N.p-P structure where N and P:repreeent thewider
: band gap semiconduqtors and.p represents the narrow band gap
semiconductor.

f) The Srip Geometry pr.ovides the confinemqrt of charges in tbe lateral , ,

direction and the loqger life for the laser diode.

t) The front and back faces of the active layer are dielectric coated so'that '

. . thereflectionsattheactivelayer-airinterfaceprovidesuflicierrtfeedback
D Thus high power output, narrow spectral width, high efficiency and high
coherence can be achieved through
.,i v the double hetero junction
,
stripe laser
Motelllzrton
I diOd6. (poctth6 gontmr errd hc.t stnk)

Mctalllzstlon
(tlegatiw contrct)
 .:Ar'*1- ,,

s
'.F

: Conrmunication
Ootical 2!
(or)
\vry
r3.(bxi) Explain the different lensing scheme availabre to,improve the .
power coupling efficiency.
Ans: (1) If the surface emifting area is larger than the fiber core area maximum
co1PIin€effioiencycanbeachievedwithoutlensingschemes.:,
(2) If surface emitting area is smaller than fibre core areas, Maximum coupling

TED

x-@
(a) Rounded pnd fibrq (b) $pherical surface of LED and
rounded end sphere
' /'

I+NA

(c) lmaging sphere (d) Non-imaging microsphere

,n
!

H Cylindrical lens (f) Taper ended fibre

,:^f
Figs.
(e)
' figure (a) shows that the emitting area is coupled with roundetl end fiber. ,

So that ftedivergediadiation from the source is focussed such that wholg .

radiation is incident fully on the fiber end surface.
O) figure (b) shows the sphdrical iwface of cE-D and rounded end or fiber. So
that the'focused radiation from the source is focussed fully on th" .orc

(c) ln figure (c), the diverged radiatioh from source is fully focussed on to the
core area.
(O In figure (d), themrcto leas, is incontactwith source and fiberzuchthatthe
, : 'emitting surface an{ thc source is in coatact with
taperended,fibre.
* G) InlEDcouplingwithfibrctogetrpatimrmrefficieney,onecanusespherical.'
surfaced LED which is in contact with spherical ended fiber (or) taper
ended fibre.
(hI ' ht hserdide coupting with fioe'^r, coupling efficiency is improved by taper
ue to their smalter
emitted area. ,
13.O)fii)Explain the fiber splicing techniques with necesssry diagrams.
a

'.
Ans: Splices are used tg mpke permanentjgint bettveen"two fibers. There are different
----- __--
types of splices.
(a) Fusion Splice: The fusion splice is carried out between two same kind of fibers
b1 heafingofthrc trvo prepared fibre ends to their fusing temperature with application
of sufficient a:<ial prcssure between two aligned fibers.

Electrode

Clamp

Advantage
(t) Low splice loss.(<
,
0.06 d$ :

(2) Perfect bonding between two fibers.

I)raw Back:
ltl uuring firsion; chemical rebctidn,may chaqge the refractive index oftwo
fILIlVDr
\

(2) Develops thermal stresses at the fusioned region.

I v-Groove spltce Here first turo fibers are placed in v-grove of glass substate.
aqe pressed together. Afterpressing aligned. The epoxy resin.the.sanlo reftaotive
of the core of the filers.
 .,r*
B.

)ntical
;;-E -
Communication

,l"i

due to slow evaporation ofepoxy.

(I )Life time ofbonding is not ascertairied
Epgxy Sin

Fig. V-grove splice

a-
Here an t"Ut ,f*n centie holc is used. Thi two fibers afe then r'
"i.rti. "
on each end for easy fiber insertion. When a fibre is inserted it expands
sther bi, plastic co(npressive force.

(l) pgrforms automatically latgral, lo+gitudinal and angular aligrrments'

(2) Elastic tube spice,loss isjn the same as value of fusion splige loss.

)..:
Since fibers are se?arate( there may be reflection loss and there isrdecrease
ccupling efficiency.
Rotary SpHce:

Single mode fihrs

 .-.+g'*.:-
B
,..,x
.t g.

26 B.E./B.Tbch. Deqree E@ition

polishing. Each:fibre is placed in each of the

;ibers are undergoing grinding and
glass rbrnile. Femrle is a precision glass capillary tube. The fibers in the'tbrule
a!€'inserted into the alignment ileeve.
(c) Mechanical Splice:

,I
.

s11*
Go:
ff
Te'

The splice is mostlyused for multimode fibers,

Each,fibre is plaod in one fomrle which is a capillary giass tube under
sion. There is a sleeve with three glass rod to get perfect alignryeltt;
aXi) A siltco1 pt-p,photodlode lm,prporated into qn o.ptical receiver
has e qurntuga- efltciengt of 60Zo at a wavelength of 09 pm. The
drrk current ls 3 nA and load reslsttnce is 4 l(fl. The incldcnt
optical power.is 200 nw and the recelver bandwidth is S,MHz.

' (1) mean s{uafe qumtnm,notse currerttr, , i

(3) mesa squrre thermal noise:current at I temperaturr: of.i0'C.

: Primary photo current, .

rr=RPo== y.po=ufL'p,
nv kc

; 9.9).(t.0.Ilo'l.cljgio+) (3* t0-? w)

[ = 1.304 W4

Mean-square Quanturn noise cturent,

./0', : 2q Io B

: = ) (1.6 x 10-re)(1,.304x I0-6X5x l0-{),

--*|.,
 tical Communication /r 't

="Q''1728 10-te A2
I '/$'' ",;

t f 0,:0.6459
vrv nA'
r

;:'
Mean-Square dar\ current,

..'

= 6'92 x lo-t' .,{

'Io?
YU

i\.
<I,*> = frln^)
-n 0.0692 fA

Mean-,Square thermal noise current,

' '
4KBT
''' <I;>r !E
= l&:)
"-'o..-
B

'I..38x 10-23 )(2.g3 Kl - . n(.

I--trt3

= 2fr21 .7 x lO-20 Az ?

. 14. (a) 1ii;lDraw and e;plain the,orperation of APD; 1, (8)

' .Ans:Itconsistsof ficur'regions.Ft: i'.-P- n*inordertodevelopaveryhigh '

electric field in ttre intinsic region as well as to impart more energy to photoelecfions
. to prorpuce.new elec, hon-holepairs by impact ionization. This irirpact ionization leads
to avalance breakdown in the rev'erse biased diode, So the APDshave high sensitivity
.- and high responsivity overp.in. diode due t9 avalanche mirltiplication..

P inpn* as shown in figure. A high resistivity P-type material is deposited as anl

. epita:riat intinsic layer on a P substr:ate; A P-t1pe diffiision or ion implarit layer is '
their made in tlre high resistivity iftinsic layer. e heavily doped z. layer is deposited

The term'heach through" arises ftom the.photodiode operation. When a'low ,

*
rwerse bias voltage is applied most of the potential drop is across the Pn jr:r.rction.

' '' J
*:-"-
-'
 .=w::''4[ill

.ri '.J*
E.

28 t
B.E./B,Tbch. Deeree Examination

I
.-t
fl: .

4.

1
Avalanche
q

Region I

l'
-----tI" :
a
l1

I .t
qf i-
.l

Fig. Strusture of RAPD and the etectric fields in it

The depletionlayerwidens witninc4asingbias until acertainvolhgeis,reached
t
1-*whifitlepd<e]MfieHaEtePn* junction is about 5 to 10 percent below
*:&at needed
to carise avalanihe breakdowu. At this point thg depletion layer jusl
t- -

['teacnes throughl'to.the nearly inhinsic region 'i' of the diode. The RApD is
tp"*,"d in the fuIly depleted mode. Light enters the diode through the ^F regior ,-
pd is absorbed in the intinsic region which also acts as the collection region for
Le photo-Eener4ted crtrri=ers. Further the electron--loole pairs are separated by the ,

ic field in the infiinsic regioll. The photogenerat6d electons drift through th*
ic region to the Pn*junction where a high electric fieldexists. trn this higl
c fieid region or avalanche region or mulliplication iegion, the charge carrier '
plicatitin takes place by impact ionization.
(i.e.,) highly accelerated elecfrons in this region collide with the bound electrons
valance band and releCIe mop pumber of elictrons as free (or) conduction

t ;.t.,
: --
;.-fo

Avalanche multiplicat ign=.ll&. - A* I

**'{z
f D]-:Total multiplied output currentlfrom diode
Ir* Primqry photocurrent. "
The value of M sub-e;greaterthan 50, fhe responsitivity ofApD is given by

R-*no
hv
M.

(or)
 . .r{
E

mqn ig s qio r,t

Qn&AL'!ie$
t4.(bxi) Discuss.the, diffqrent noise sources and,disturbances r" jl1
,]..opticrlpulsedetectibnpecha3ism.,-.-.
Ans: Dlfferent Noises: There are so many detectii noises' ' ' ' .

I ::
i) Quantum or shot noise: fhis arises from the Statistical nature of the producrior

'
Meansquarequantumnpisecurrentisgivenby: ' "' :' '

where Ii= Primary Photo current '

:

S = Bandwidth of photo detector

lo rtv!=Avalanche multipiication factor.
!

'fM) = Noise [gu{e ariqed with the randem nature of the avalanchip.rocesg.
ll) Dark current nirlse: lttris arises due to the dark cW,ett which is the currelr'
flow tluough the bias,circuit evpnthpugh there ig no inCidenr It is divided intt
-tight'
two types of noises.

' ii)
- :.:
surface &ark curreirt hoise (or) surface leakage cument noise.

Mean square of the surface dark current noise, Psd = \qlp'

"' ' t 't '
in'=io'+"f+i*"
'N ?o 'u 'il' -

Johnson Noise: Thi photoTdetector load resisto,.l who{ variation wittr temperatuff
gives thermal nois6 or Jorihson noise. Mean square of the thgrmql iroise current ir
givenby,

s. t3
Thus izW +ilsd +i2r+{ '
 - .T4""ir'''
"i- ,

g.

:
,.30 B.E./B.Tech. Deprce Exariina.ioir ,$

14.(Oxlt)Dertve.an'eipresslon for thb bit error rate of an optical digilut

r recgiver. i
-_ (I1!)
Ans:, Probablltty of Err{ri ltlis w,ell knorytkat the noise intqrferencc ieadi t,,
,
i wro,ng decision at the receiver. Proliability 6f error.'Pe' is a good rneasur.; for
performance of the detector:

FIg.
Figur-e shows that original signal pulse'taln tansinitted is givenby (l0l), But
. due to noise, the noise voltage goes above the threshold leriding to wrong decision
bytherecbrvelandthepulsetrainouSuttromreceiverisgivenby(111). Probability .
ofenor means that a transmitted ' I' is misinterpreted as a '0' (or) transmitted '0'
is misinterpreted as a'l' by the receiver. This is called.as bit error rate. @ER) ,

N" il'
'r Frrr
,,'rtr-'ry,
BER=; -
Brt

I
= ,Tb is th1 number of bits,per second. Fgr elanrple the protlability of .

eror or BER equits to'l0l?, thtn the ayerage, one emor occurs for every giga pulses

Lrrrn)dY
which means that the probability that the equalizer output voltages is less tan than
'V' when a ' 1' pulse was sent, and -

fr(,') = f ot, to\;dY ,

which means that the probabilitv thit the equalizer ou-tput voltage i9 greater than .1 P.
-'Ihenas '0' was transmitted. Ifttre threshold volage is 'Vth' then the errorproblbility
 -. i?.+$x-_ *:.-
f.
'
''
'i+
g.

tical Gommunication' ;f

Pe.J r a P,(V;,,) + b Pr(V,o).

\

:
For unbiased data with equal probability of I and 0 occurrences, 4 = D 0.5, _
let trs assume that ttre noise has a gaussian.probabiiity density functiOn with 7e, fO
\

f(nldn=+ e-n'rz&4n
{2nc'
wherr.o2 in the noise variance andfln) is the probability density function. Let all" l
pulse have a multitude '12 volt.

f I ,-r,rn
rzczd.y,o

*n r" subscrip.'0' denotes the presence of a '0t bits. Simply the probability of
error in the care of ' I ' pulse is the probability; that the sampled signal plus noise falls ''
below Vn. That is,
r 12

,, r Pr(Yril= I

.r'l
=@Le "'' o.r,^
:
.?
Thepobability,of orror 'Fai is the decoding of any digit (0 or 1). ''

. ,

j- ' "'
\ ,

: 2$
,h"or, I
edX=
GI
e-t'.dY
..:
iff t l Explain the signilicance of link priwer budget and rise time budget
t' with one illustration foi each. (16)
t.fup Link Power Bqdget: The main aim of power budget is to have enougfr
- :^rhcr at the receiver so as to maintain reliable performance during the life time of
,,.'....--
'32- . : '8.E./B.Tbth.DegreeExaminatior.
the entire fiber optic system. The mininiurn poWer required at the receiver is 'P*.'
Thcaveragg power"launched,at the trar.rsmrtter is 'P,'.
b
.:. Pu= P** Pr* Pru - '.----'

PsN-System Margln.
The sy.stem margin is the link power margin which is normally added with th
total power loss. It is included in the analysis -of power budget. Temperatur
'fluchntions
and,c,oinpqnentdegradation and its value ilahgut 6 dP':, . ,
Thus total power loss includes cable attenualion loss,tonnector loss an
splice lgss,

where ooo =
"u*1rul
loss = 9F * ^r= oonnector loss arrd

,
a
--sP = Solice loss.
' -r---- --
irtfr"tiptassion an P*,, andP,oare the optical poweravirilableatthe outp
.end of the iiber of length '/i The required receiver sensitivity is42 dB.
'\
For Examplo: Ta[ing GaAstdiode laser (]', = 0.85 pm and Power = I mW), receiv
requires about 5Q00 photons/bit to:operate in a reliabte manner with a BER belc
l0e since for B = 50 Mb/s.
l

P : Np
Prn:
:

iV hY
hVRB -
\

t-

7\ r' \' -- -
0-85 x i0-6
,:t
fD

t-tt:'

t'

P.-P
L't trec -P -0+42d8-6dB
r -P sT

:36 dB
/

.'. Channel loss, Pr:36 dB

.It is estimated that 0r= 3.5 dB/Ih
cr,c+or*l1+ l:2dB
'e,L=86-Zi34dB
 ', 33
tical'Cotnmun
'
The' maximum ffansmission distancer
. i. '
i

r34
t 34
!L'= =- =9.7 km
o"
-'I J.)

laser diode at0.85 lun wavelength th3 maxim,m

lengh 6f
Thus with I mW
'tb.t';;. ili-and requirea receiver sensitivity 5000 ph-o.tonstbit' 'Economic
based transmitters'
oiria.rrtio* witl gi#the riglt chofce of LASERoTLBD,
powef of 50 pw (-
In the case ofGaAlAs LED which can give an optical
I an*l ilt-;Jmum value of channel loss is only 29 {B'
6 Km with ctr= 3.5 dB Km
.;. Transmission distance is reducid to
The dispersion equalization penaltyQ will increase in optical polrer at the

eceiver. For Gaussian shaped pulses, :

D,=F)U
vhere,-r, * y. full width pulse brodening due to dispersion on the lirft'
! - git interval or Period'
t

n!

Thus
Factor 2 appears to account for the gap between bits.

,i

It is to ensure that the system is a6le tq operate properly at

the irrtended bit .

time uraget an*vsis is a methpd for detenirining tlre

dispersion limitation
, ,.r*. nir. "

Thg risg time.'f,' of a linear systcm is defined as the tinre iluring'which the i

'*tp"G in *.. dom 109o to,o6z1 ortu. ui*."P*.:}t: '

. ,' , . ' ,.
,There is an inverse relationship betrnleen the bandydt!.r
:n: 91yfm1m'it
time of the simple tRc
rate 'Br' and the rise titne t}by conqideration of
the nse

.*
 .g,Ei*"'*i-
B
. .Jx
s.

(d) Transfer funstion H(w)

For step inputvohg*, the oututvoltage fromthe Rc filtei, Circuit is given by,

':'(A)

T-= 0n 9) Rc = 2.2 Rc ... (B)

The I(w) of the circuit is obtained by taking the Fourier transform.

.

The electical band width for this circuit atwhich lH(w)l is, given by,
 , r,=**4.J'*iffiE*r,
__t'15
g.

}nRC
Using (B) we get --
1l

.i
i,
t
2.2 0.35 \
=-
.l
=-
ZnB B
It iscornmon to use T,B = 0.35 in the design of opical fiber communicatior
svstem
-J ----;

0.35
Br(ma:r) =+lr
in IrIRZ fofnrat bandwidth is half the valueof maximum birate.
'But

lf

i.e.,{s {'a35/Br,;;forMformat ' i

{0.7018n*, for NRZ fonnat '

ADVAI\TAGES OT NRZ CODING:

o Minimumbandwidthisneeded, : '

DISADVADEAGEOFI{RZCODING:,
o Average power input tgthe receiver is. dependent on data pattems.
. At trigtr 16*l qf receivedpowerit giveBASE lnfeW4ryOEREFFECT.

'o hovidestimirg'infonrrhiion.
-.

DISADVAIYTAGE OF 4 CODE:'
,,1
. , It requires high bandwidth since each data bit is encoded as two optical line
code bits. \\:--
-\-\

'15.(bXi)Explaintheprincipleofoper,ationofErbirimdopedIiberampliIier.
(10) s
Ans: Frsic Principles of SONET/SDrI Networks: The lack of an intemational,
standard in the teleconrmunication mdusty duting !e!O-s led to the advent of a new
 Examination
standard, first called .the synchronous optical network (SONET) ?{d later'termed
the synchronous digital hierarchy (SDH).
SONET is a syrrchronous:optical netrvork which has'a master.clock that
controls the timing of events all through the opfical fibernetwork. sONET is the
wide area network operating at a base rate of 155.52 M bits/"with expansionn
capability to achieve data rates of several G bity,. soNET is also considered by
the cITT (International conslltative.committee on relegraphy andrelephony) as
an intemational standarrd andthis international version is called Synchronous Oigitat

I SONET defines a synchronous frame stnrcture for sending the multiplexed

t
digital haffic and the operations procedures over optical fibres. The basic building
F
r block of the soNET has a bit rate of 51.8a Mb/s. The correspor,rding opticJ
si$al is refemed to as oc - I where oc stands for t'optical carriet''. The,basic
building block of the sDH has a bit rate of 155.52 vrys *o is referred to as
. STM-I where STM stands for a synchronous transport module. A useful feature
of SONET and SDH is that higher levels have a bit rate that is an exact multiple
of the basic bit rate.
Table: SONETTSDH Bit Rates:
STAI,\TDARD.SONEI STAI\IDARDSDH. LINEBATE CHANNELS
t

LEVEL BT Mb/s
,oc-I -, 51.84 672
oc-3 STM,- I 155.52
,.
2Arc
oc -12 STM -4 622.08
oc-4g STM. 16 2,+99.32 322s6'
oc - lg2 STM - 64 9953.29 '129;024
The table shows tle corrrspondonce betwqen SONET andTDtr for a
different bit rates. The SDII provides an international standard that appears to be
, well adopted. srM-64 systems'operati,4g near locB/s is the latest version. As
,'shoum in the Table, for each level the b]trate is increased. Simultaneously the
number of chanels are alss increased. So. tansmission and rebeption of signals
can be done at avery fast manner. Recently a transport networkis afr. *arlrru

, network (wAN) called a multiwavelength optical net#ork (MONET) is designed

to incorporate the existing diverse switching technologies (SoNET or SDH) into
a longhaul all optical network (AoN) by using the wDM technotogy.
: clearly'the wDM technique has revolutioned the state"ror, "o**t
of the art of light wave

t5. (b) (ti) Give a brief'accfUt of th6 principtes of SOiyET. (6)

i
t
3
Eff,FsEs
-:.=r6+..*@f,6,
.F
H}

Orticel€osrmunicetioL , 37

--
ANS: ERBIUM DOPED AMPLIFIED:
the- repeatert
In the .long distance optical fiber communication system' Tt .

KMi Previo.sly it yas tr conversion

.i
"";;il;;;;iairt*oi"ro"t
i 00
by electical amplifiers
!w '
"
of ontical energv into elcctical energy and amplification
ilIilenrt;;niersion of erecrical *rgy
into optical energv. Such methods 1ot
;&ffi;;;; ;;.r.. and cornplex!,, ai sical corrmunicalion svstem but
9.,
also- reduce the operational bandwidth of the system
the
', f,ntiothCre is alarge neduction is the cost 6fthe laying a1lmaintaining
optiffifril* i"*-rr*pr;68, *ht h is in the form of fibgr coupler, and
"
easy fabrication.
stimulated Raman '
Initially optioal anrplifiers we,re designed utilizing coupler
:
:,
*-
d#ilil;;il"# scattering. Bv rnjecting hi4 power lTt'.btT T^1
" l0 km) ora -
p,"ttpitie-;A*t*intoaundopdrclatively long length fibef
ir
*l (
t.
loeed
(-
l0 m) ilre laser action i.s.actriew$ttro;1$ the non.linear
F
:
E
,norir.i*u fiber
ffi;-dd*;a*iaiotn"nu.randoanpqovideontical amr]ificlion r.-
f-
&

Mianwhilc mollerauar dCIigpd an obium doped silica

fi\r laser alnlifier'
It il ils[---A;.t* in a suoJtength fib9rr! *!t'q.q".ryd si-lica lP",Pt
iOiuffi ili.*t tioo isabout I6s 4o nslm'. ror an optical slgnal arplifcation
it a waveleneth of
at wavelerrgth or r.iir* il
Ctetp hser diode qperating
1.48 pm is used as a pumping source'
. In erbium doei:silica nU"r, onty,the tansition betrreen
E, and Er-is 10070
*di;;;;;;;,.ir'o,no ttansistions are non-radiative. Due to the rapid
;;;it*rt*-;rre; uo in the ener$r burd E , evel llous\tne lbt:T:io"
is at I :55 fq frol thb bottom
;ilffi is at l.a8 lrm the emission wavelengt]rpYP.power
' of the enlrgy Ut"a f;, Hence at some tnresUop
level' poPulation
energy lt".tt by
inversion may qrist. dtirnogti,oo ofthe erbium ions in the 3, Pt
1,
i
;#;G;dh"d,* eir/e:leser action suctr that emittd photon and stimulating
le
.Thusthe-'*.,isamplifigd.qne"rue.sierr*[av.ets!3lvntneoCIicalfiber

ain'lifier, the pu*p po*i* Ue er;dua[V Aec-reasea since the signal i:yltifi 1
to set
#;;"b;;ffi, rUh there is an-optinirum l9nqlh for the fiper amptifter
-.raaximumgain.
t ..- --l
Pump{ng Sourea
\x 1'48Pm\
--f> ) -nmilified Signal
Weak Sbnal C
,'*L = 1.55 pm
L = 1.55 pm
Fig. Optical Fiber LaeerAmplifler
' '-/
F

r
r 3
F
t
.
It,shows the opticai frber laser anr!,lifier which is in the fofm gf aT goupler
There are two short fibers whosg middle portions are welded together. ';
Advantagesof EDFA'I:.
I ,,
They cqfr work efliciently eve.n at higlrer bit rates' ,

2: They have aiarge rpr.t rt bandwidth of 4orun: so'that 1

qingle amplifie

3; ' Ttrey have high gain when compared with semiconductor laser amplifio
and Raman amPlifiers.
4. Thesecanbe onlyinsotedinthis Iiberoptic linkwithminimal insertionlos

5: . Tlie noise ihtroduced by fiber srhplifien is very low andis about 3 dB.
Dlsadvantages of EDF A's:
I . When we usb the EDFA for a w3velength diVision multiplexed signal, ne'
'ftequ.ncies are developed- l\f*fzaldd, ue signqf frigufncigs' then.t\r
*.y ur4 * f, * fllf any of these frequencie,s coinlides with an existin
'siglrat *lqrlir.lt then it will resylt in cross talk between the differe!

'
2. EDFA's are fimited to use at 1550 nm system only. For 1300 nm system
oplical fiber implifiers dop{ with praseodymium are under derelopmen
f. Ttrey require ligh:pump power in the range SO-IO0 illW'
4. ilhese can not beuied' iu the fiber hpving very short lengths:

!f
 -,rsr.*@f,6,
',s
F B,

:' v4159
CI
;
s
;!

.a
.!i

B.E./B.Tech.'DEGREE EXAMINATION, APRILIMAY 2009.

l

'i

(Regulation 2004)

.
'(Common
Ec 1492-oprrcAl co.MruU,utcan0ru
to B.E, (Part-Time) Stxth Setnester Regulation 2O05)
Ttml : fhree houns Maximum : lfi) marks
:,, Answer ALL questions.
PART A-{10 x'2 = 20 marks)
1. ' A silica optical liber wlth a core diemeter large-enough to be
considered by ray theory analysis has a core refraetive index of
1.50 and a cladding refractive index of 1.47. Determine the numerical
aperture and acceptance angle in air for the liber.
Ans: Numerical aperture (N.A) = (nr' - nr')"

Mention the advantages gnd disadvantages of monomode fiber over

Ans:
q:r.. Advrntages:
Aur-rrsEer.

Since the information transmission capacity in optical fiber is inversly

poportional to dispersion, the mondmode fiLers aremore zuitrible for long diitance

dvantages: .

Launching of lighi into mbnornode fibers and,jointing of two fibers are very
difficult.
Fabrication isvery difficult end so the fiber is so costly.
% What is intermodal dispersion?? What is it cause? ..4

.lt
1i. Ans: Intermodal dispersion arises due to the variation of group velocity for
:ilr$ mode ata
t a single frequeqcy. Different modes arrive at the exit end of the
ilbcr at different tirnes. So there is multimode dispersion and hence there is
of the signal pulses.
Iryening
 - :'wITaEIUHT:
.,lt
B.
*
4. $stinguish intrinsic and extrinsic abporption.
ians: Intrinsic Extrinsic
* Associated with SiOz 'tG Results ftom transition metal

.a'
t:

{r Infrared dbsorption by Si
- 0 -,
coupling
* UV absorption is due to * Also results from OH ions, I
:

electronic absorption bands.

5. What iC populatlon lrtverslon?
..
Ans:. Underthermal e4rilibrium, the lowereaerg! levelJ, of the two level
fflo.i" systemcontainsmoreatoms thanupperenergy levelEl. To achieveoptical
$mptfication it is must to create non-equilibrium distibutions of atoms srrch that
energy leVel i.e.,
$po,pulation of the upper energy level is greater than ofthe lower
> If,. This condition is knowr as population inversion.
-t
$4
Compare LED source and ILD source.
t
.a Ans: tED ILD .:
^ 9'.-
,i3
*qf
t'
t, * 4

Incoherent * Coherent ,

d'
li
!,
{,
* For multimode fibers ody :
* For multi and single mode
"a

r,*
"&
fibers
#
5lt 1

*.!
* Modulation Bandwidth : 100 MHz * 10 GHz
i,ST
* Large beam divergence due to
!r Low beam divergence due to
' spontaneous einission stimulated emission.
Ga As has a band gap eenrgy of 1.43 ey at 300K. Determine the
navelength above whiEh an intrinsic photodetector fabricated from
o
this m&terial witl cease to operate. ,
;

Deline the quantum efficiency and re$ponsivity of a photodetectgr.

\ ..
Ans: Quantum Efliciency:
It is defrned as the fraction of incident photons'which are absorbed by the
detector and ge_frcrated elecffon which are collected at the detector terminal.
 3*

--
tst=
number of electrorts iotletcted
'
I|
numher of incideit photons

ResponsivltY:
It is defined as the ratig of nhoji current to the incident photon. .

gl'whatiethepurpoieofitse.itmebudgetanalyels?
Ansl Rise time budget is io ensure that the system is able to operate ptoperly
* tfr. inioatd bit rate. Fise time budge analysis is a method for determining the
dispersion liinitation of an optical fiber link. The colgePt o{ri1e1ime isused tc
atloo,qte the bandwidth among various components ofthe topical fiber
finls.
10. What ls EDFA? . l
Ans: It is optic fiber doped with'erbium. .

Erbium has soine interestingppperties for optica! communications, Photons

at 1480 and 980 nrn'wavelenglhs activate the electons into metastablb state.
Electrons falling back emit light at 1550 nnr hence it is a low loss wavelbngth
region for silica optical fibers. It amplifies a signal by stimulated emission. Ilere
thJ hsersignal lies at the waveiength of 1550 nm. The noise signal at )' = 936
'

d".tq: [.^stimulated
emission ly means of electons falling back to meta stable
state at 1480 mm.

1r. (a) Derive an expression for numDer of modes propageting in a

Ans: In:the ggaded index fitrers, refi'active,index is lrotconstant but decreases

'
with radial distance from'a maximum value ofn 1 at the a:cis to atonstant rralue n"
in *re cladding. The index variatioa is represented as

n,[l-2A]"'=nr(l-A) =nt for rza

The refractive index variation in the core obey$ the powirs law relationship. .
'Where
A is the relative iefractive index differerrpt. " 6 '? lives the characteristic
. refractive index profrlqofthe fiber core.
When cl e [, the refractive index profile is tiartgular I

Wtren o = 2, the refractive in'Jex profile is parabolic :

 #
Tech. Examination (A 2008)

when 6s. = oo , the refractivb index is constant and it is a step index ptofile'

'2

Rnlrlrtivn rndcx la(rl}

,, {, Srdirl dhtrnct trl

* | Corl
i
" *ti*

Flg. t Posrlbh.flber nfnc{lve lndefi proflleo lor dltbnnt vrlueo of cl

' Figurg 1 illuslrates possible filel refractie index profiles fordiflbr-ent vales of
"
. " 4'", f.rt6olic reftactive index ppfiles p*a.o.. the best results for
multimode
optical propagation. The index difference 6 for the graded'index fibe.riS given

The re'fractive index profile and ray tansmission of amultimode graded index
fiber is shown in Figure 2

F|g.2Therefnctlvelndexprofllerndrayttansmlrrlonlna
' multlmode gradad lndex flber
It is clear from the.Iigrue that the meridional rays follow curved paths through
the fiber coret The gfadual decrease in refi'active indp-x from the cenEe fiOm the

The refiiaction sfthe rays takes plnce as it inci<lent on a iarge numberof high
to lo.w indexinterfaces. Whenthe ral ingidgnt angle becorres greater than the
critital angle then the total internal reflection takes filace and the ray havels back
 s
tieal Communiiation (EC- t4}z
t&arOs the core axis,'because gf refraction. In expanded ray diagran that shows
refraction at the various high to low index interface a with in a graded index fiber
ibihown in figure
: 3.
I.llrl tnternrl reflet'tion

n-
R;
Ilr
flt
n..

=a

,t

3 An expanded
Fig.
' ray dlagram shrowing refraction at the vadous high to low lndex
lnterfaces within a graded tndex flber, giving an overalt curved ray path
Numerical aperture for e graded index fiber: For a graded index fiber,
the numerical aperture (ff2{) is a function of position across the core end face.
The light incident on the fibre core at a point "/'will propagates as a guided mode
ifandonlyifit is incidentwithingthelocal numerical aperture. Atthatpointwhere
te local numerical aperture is defined as.

o^t.4(0) ,-[*I ftr r<a . ,. (3)

0 forr>a
n, is &e,cladding refractive index, and "q " is the core radius and

-The axial numerical hpertuE is same as the numerical aperture for step-

l/l(0) = [r2(0) - nr'l'o n;)': = r, .li ...(4)

=i(n: =
-I ..

As 'r' moves from the core a:riq to the irunundary (core-cladding),'the numerical
for graded index fiber decreases from N/(0) to zero.
of modes in graded index fiber:
The total number of guided modes or rirode yolume "M;" sipported by the
index fiber is
 M=
I

.Ln .

,IL
Where K= ? - propagation constant
"

The normalised freuqency "Iz' for the fiber when 6 << I is given by
. , Y-nrKa(26)trz
Substituting equation (6) in (5) gives

foria parabolic refractive index profile fiber core ( a = 2) &/, is given by

.Y2
24
11. (b) (i) Derive an expression for numerical aperature of a step Indei
fibre. (10)
.l Ansl Figure I illustates, launching of a light ray into the fiber core. "A'.'is
the meridional ray which makes an angle g. within the core cladding interface.
'i
r Thi6 ray enters the fiber sore at an angle 0 to the fiber axis and,it gets refracted
"
i,i
i.l
at the air-core interface. Ttre reflected light reaches the core-cladding interface

Ii. at a critical angle 0. (minimum angle of

incidence that'causes total intijrnal
I reflection) get propagated down the fiber length. Figure I also illustatps prop4gation
of another light ray "B" which is launched to fiber at an angle greater *xn 0..
I
f This ray B.gets refracted into the cladding and get lost by radiatioar. For the rays
f, to propagate along the frbenby total internal refraction, the light ray must be
incident on the fiber core within an:acceptpnce cone defined by the conical half
E.
!fo *** 0 . 0 is the maximum acceptarice an-lle at which thb fight enters the fiber
" "
I u cote from launching medilinl 0. is also referredto as the total acceptance angle.
1 0. (acceptance angle) mustbe less than the.critical angle at the fibercore -
larmchilg medium interface otherwise lightwill be totally reflected back tothe
): larmching medium. The acceptance angle can be derived from figure 2.

i-
 ,,
's,

-L402
L
calCommg4aqou

-t

\,'.
\rl
./

launchlng $ght into at n optical fiber

Frg. t The acceptance angro b., wlren
'i{.
I

,.'
n

lry:l:l]l}-ll.ilgltJl ' '' r rD opthal fiber in air at an inPut

2 rhe ray path for a meridional rav
\ . {lgles le,"ahan thelCceptance'angle lor tho
lbq ,,. ,.r. . , . ,
to the fiber axis, This *91.'
The ti e fiber core at an angre 0 ,

*rl"ilffi;Loiil""5"i"rortrerruel'*"ryY,:l':*If
into thE core'-According to Slell'
#il,f;ffi;ffi;;"rc*nitua
s Law ;t
Refraction at the core-air interface.
... (1)
'mosin0.=qsine2
it U2
the
in
otal internal reflection takes place. Becaule
GraqsrrE intbrface$otal
tnrE itaAAing
At Coie "'t"^;;;fr;t *;;;fn;e is greater than the critical angle and
e ofincidence at corcl

ir;;;;iJ q. The equation (r) can also be written as'

.s \
... (2) :

,.. (3)'
a

'
-O--sin Q,
no r I = nrI cos 0

t-sinz$
f<no1 case for total internal reflectior
0 . (whi9h
to is f
qtinS
$ Q is'eqtral .1s
can be qritten a:
Ol is'aquA to 0 ., then equation CI)

..rL.
 ,rffil*iHtrUm,
*
.F

mo,Sin 0, = flt

zo
a'

stn 0. = fl, ... (5)

,rosin 0.* ,{m i.. (6)

t Equation (6) shows the relation between acceptance angle and the refractive
i ' inaices of tt ree mediums (air (z) core
(n,), claddir-rg (rr), The [ght rays which is
t
'
i"*[.J *ithin the
"...pi*".
[oo. *itt'ue nappeo or gpided into the fiber ctire
i.e., the meridional rays orrer the;ange 0
'0 g (
0. will be pfopagated within
,

. NUrnerical aperlqg is an important optical parameter. It is oftenreferred as

1 figure of merit of tne fiUer. it says about maximum
light gathering capacity of thc
ll nU.r. Equation (6) is, $e actual formula to calcrrlatg nu,merical aperture (NA).
l

,.. (7)
I

ia=nosin o.= ^{f'A

Generally the light is launched to the fiber cote from air mediurtl so equation
I
,

-.||reducds to
(7)
rl
i

.,, (8)

, t frCI= I for air]

The rnunerical aferturg'alsorbe diven in terms of relative refractive index '

difftrences 6 between core and cladding. .,

' \

''r
"?
- ni ... (9)

frr*frz
AF,,t r.l
for A <<1

I
A.- relati'ie refractive index difference

,i

I
r
n
i - cladding refractive index.
i
,
I
I
f
 Comm,r.ic@
i i"*ion (g) can be written interm$ of relative reftactive indbx differencq 4,,

NA = sin e'a- =

sin 0'. =
W
(3' + n)(nr
( 10)

sin'Q. = 2nr(n, f '.' flt = nrl

[...ory1
lin. o. = L n,J
... .
i- .

* ... (11)
NA=sin 0.' ^{ffi or ntJzt
a

From equQtion (1f)'if the numerical aperture is knourn then maxlmum

\

acceptance **i,.:::Tived as follows' :

g"=sin'r (Nr{) "' (13) .'

the core axis

u. \:' fiil
--- ft) \' The relative refractive index difference between
when the-
and the cladding of a- graded index 1lbre is 0J7o
values for
'

refractive.index at the cote I axis is 1'45; Esdnate

ln" numericsl aperature of the libre along the axip when the

Ans: Numerical APerture (NA) = nr(Z A)'n

12.(q)withtheeidofdiagramsdiscussthevariouslossesoccuring

12. (b)
- A 5 km optical link consists of m-ultimoae f9 -dT n?.tt ''
, with " .oi. refractive index of 1.5 and 3 relative refrlciive ,
'.;..lndexdifferenceoflYo.Estimrtelhe!:t",diffcrencebetween

poi"" Utga.dening due to intermodal disperslon 9n lhe link

 .F
E

Expression

Fig. t The paths taken by the axiat and an extreme meiidional ray
ti in a perfeh muttimode step indpx liber
Intermodal dispersion in multimode step index frber is estimated by finding
i the delay between the faster and slowest modes propagating in the step index'
I frber. The paths.taken by axial ray and
-
extem" *rridiorrlf ray-is sho*n in nguie.
Ft Thu delay differenci is directly related to their path length bicause the light rays
i are travailirtg at the sane velocity within the constant refractive index core. The
i time taken by the axial ray to travel along the fiber is given by

\
T=-=-
*min VelOCity CI ,nr' C
...(l)
Where n, is the core refractive index. The exheme meridional ray exhibits
morimum delay time T^*where
't'*

rF
I/cos0 -YEv
Ln,
-:-l
' max -t -
c I nL Ccos0
i

By snell's law of refraction at the core-cladding int6rface.'

rT
n2
sin tp,c : e ... (3)
n:cos 1

Hence
tj

T-Lni
m(N Cn,

The delay difference *6 1': between the axial ray artd extreme meridional
ray is obtained by

ui
 -rc-"*@*F

*
..*

calCommunig@
L\t | !"ry!'
64=?iorT T*ro- cttz c

a\T)
a

Ln,
cflz
' : --'
be
is the refractive indice difference and equation (01 can also
written

Ln,
/\
l ftt_flt I LnrA
or,- c t n2 ) c

and by substituting for A then s 4 can be written as

I

ors_ Znrc
,

pulse
Tf,eapproximate expressions fordelay difference used to estimate.rqrs
broadening.
area as shovrn in
., when the optical input to the fiber is a pulsepdr) ofq$t
 .F

Examination S'

Where
Ifit is the mean value of the p-ulse.
t

! s
k
r
I

:-

E
, The,mCan value of the:input pulse is eero:hence:

F
E
r'
F

Mr*

t [r'llt" t(ors\2
- 6rs[sj_r",, 3( z )
-l
-
I 3-l I

r(N/)2
G**?ffi A=

Thg,equation (t$ gives the rms pulse response of a m-rrltimodg step,io.lq

fiber in which intennodal dillpersion dominates. The pulse broadening ir Oir.tf
propo:tioiial to indice itifference and the fiber length. Reductiori in " A " reduct
the acceptance angle and thereby numerical aperture and hence launching becomt

Problem:Thedelriydifferenccisgivenby,.....).,.
' Ln,A,
or-r: _c

2*, 1n3.-t
3xI03 (
xlJx0.01
-'-,;- :---
=* o,;.,t, . = I.5ng.
3xlOE
 (.rt
S.

tical Commu
ttt* rms pulse broadening due to intermodal dispersion is,

LnrL,
or=ffi 1

3x103 i1.5x0.01
: 0.43 ns.
'3 * l08

n. (a) Explain brielly the three key involved in the laser action
proc-g3s9s
: ' 'p...iibe
for t"nry,perot resonator laser diode, modes ud
"
rthreshold:conditions. Obtaln its rate cquaiions for stqafy stete
'
output.
\ns: Key Process:
The term LASER is.an acronlmt for (Light Amplifrcation by Stimulated
lmission ofnaaiation). Laser action nas leen obtained using may differerri materials
yluding sarys f.iq.u n*n orcarbondioxile,liquids
*U t"rt*,H1ff#fi;;
fne s"micbnductor laser uses the solid semiconductor as the lasiir
,il;-irbrt r),f,r*r the laser sornces used are atmost exclusive]1 sTirco:du*
aier diodes. ihe output radiation is highly monochromatic and the light beam is

Principle of oPeration:
Laser action is the rezult of three key processes
(i) Fhoton.absorpdon (ii) spontaneous ernission {ili) stimulated emission'

Figure (1) iltustates a two energy level or state atomic system But:r
is initiatiy in the lower energy state 8,, When.Photon with energy Er: E,
'rtom is
"
Enciaent on the atom it may be eicited into the higher energy state E, (Excited'
refered to, as
) through ibsqrptiog of phOton. Ttiis procqss 'is:sorqetimeb
"rrt
, dmulated
.rake
tUotpdon. When atom is $tiaffv in the higher:"rjgv st1te r',it c1
a fansition from excited tt"t. 8, to lower energy statg E, and emission
of
fioton takes place. This emission process can occur in two ways.

(ii) Stimulatedemission.'
lpontaneous emission:
-
Atoms in the excited'state E, return back to the lower energy state in an
:ntirely random maflner.
Sdmuleted emlssion:
When a ptoton of energy equal to the enery difference between two states
 : ,.f ^S.

l{ , l' . B.E./B.Tbch. Deeree Examination (Apr./lvlat 206

- -
(4- Er)interacts with the atom in the upper energy state and makes it to retun

lnitret rtah: fiiraf $-l,ttr."

*r*=*---..-'.f --*.F_e

.\tlorptilrn

ul

r
,*
S6rnntane.bus
ernission
Mrfu-

$,,*-*---L*
(t))

$timulatsd .nfrnaltn"f-
0mtrgign
.a
I
'l
**-trcrt$rrcr r.,r*rr m

{Ci

FE, I Energy **edhgtrm rhorm : (r) $sorptlon; (b)spontaneour emligibn;

(c) ltimutatad emission. nre ltacl cot
!$\{o s: rtate of the arom beforo and after a
transhlon takcs place
The spontaneous and stimulated emission process are lllustate{ in frgur,
tO) and (c) respectively. Spontmeous emissionprocess is ratrdom in natroelan
the light emitted is incoherent. This is the basic mechanismfor light generatior

The stimulated esrissior process gives the laser its special properties as al
optical sources. Photon produced by stimulated emission is generally of identica
enelgy to the one which caused it. A photon of encrgy 'hv! will not necessarilj
alwaysstimulate anotherlihoton withenerry'by' photons may ue stimulated ove
a small range.of energies around 'hv? providing an ernission which has a initr
tequency or wavele,ngth spread (linemridttr). The light associated with the stimulater
and stimulating photon is in phase and has the same polarization. i.e., the ligh
' radiation due to spontaneous emission is'coherent in nature.
Coherentmgan, when an atpm is stimulated to emit light eaerry by an inciden
wave, the liberated energyca be added to the wave irl ionstnrctive manner an(
pmovides amplifi cation.
 t*
15
,ptgtl Ssrnrn+$ir.,qtipE, (.p$. 1491) -..- .-

abay Perot Resorter: ...

.rserDiodeModesandThreshoIdCon.dltlpn$:..
Semiconductor laser diodes are prepared over LBD for the oRtical_ tlf
rmmunicatio+ systems requinng baildwidth gr-eater than approxim$elv
lmtvff1?

(ii) optical bandwidths of 2nm.

(iii) high corpling efficiencY.
Virualy all laser diodes in use are multilayered heterojuncdon devicCs' Ibe
.?
onstuctionbf hser diode is morc compl[carcd because of additional requirement
f current confinement in a small lasing cavify, The stimulated emission in
erniconductorlasers arises ftonr optical tansitions bet rcen distibutions ofenergy
iate in the valence ond condrrction bands. Stimulated.cmission via populatiort
rversion is achievedin an infsic semiconductor by the injection of electnons into
p conduction band ofthe injection ofelechons inlo the conduction band ofthe
raterial. This is illustated in figure 2.
rigure-2(a) shows the situation at absolute,zero. when the conduction band
, on*irio;;*;r;el".oonr injected into
the material fill the lower energy
ibtr"io tU" copductign band gap upio the injection :nerqy o-r quasi {elailwel for
qqual
I tcctons. Sinea,thq qharge neurdity is iqnscrved with the.rnaterial-an

, e".ity of holes is created

in the top of the valence band by the absence'of
! bctrons as shown in figqrq 2(b): , - , ,

Ilf,ffil rirt*t(t
rlttstt(|n
l\sJ.rji.l
tltr&.r.r. t?tt$t.

V*l+r",'tr*
Ir*ntS

Eg. 2 The filled electron states for an lntrlnglc dlrcct bandga? :.Tjcondudor at
absolute eero: (al equlliHum; {b} wtth"hlgh,canlar infectlon
.
 :..:i.: - : : - ;::' '. I \t-1,ua,, .,
flg. 3'FeUry-P.erC rcsod*or crvlty for'a lass dloda,The,cloard gqdt fuiletlon as
partialty reflsting mlrmrc. The unused end (the rear facet) can bs co*ted wlth a
the,cavitf [etg th* th. llg { beam omerging
dlalestrlo, rcffrrtor to,rcducr,oFlfc{,losr"ln
from tha hrer lormg a ver{oq! effipqA eve.n thglg!-the lasing spOt at the a$tu0'8r01
b
F:
* ' - tacst ls a horlzontrl elllpe :

I Sinco the necessary conduction lrand states are occupied, the incident photons
with energy '?r' but less than the separation energy of the quasi fermilevels E, =
8",'f^.i414otbeotserbed. But, thesbphotons indrrce adownwardtansitio.nof
kFi'
s'
an clqct-oa ftom the lilled conduction band states' into empty valence band states.
r: .
The basic condition for stimulated emission enerry as weli as the band gap enerry
and it may be defined as

largelymaihtained. . ,,., ,.. :. ,,,,.

The radiation in the laser diode is generated r{ithin a fabry-perot resonator

This cavity is of 250 to 500 F m long, 5 - I 5 P m wide and 0. I - 0.2 [ m thick.

Th*e tbteo dimen'sionsare getreraily refbned to as thb longilu$nal, lateral and
tranwerse dimensions ofthe cavity. The cavity is enclosed by a pair of flat, partially
reflecting mirrors. They are facing each other. The minor facets are constructed
by makingfte two @lel cteffs along natural bleavage planes gftr Vtnlconductor
crystal the piryosc of this mirror is to p_rovide optical feedback in the longitudinal
direction. The device is converted into an oscillator with gain mechanisms that
compensates for the optical losses in the cavity.
The major losses result from factors such as absorptions and scattering in the
amplifring meditrm. Absorption, scattering and diffraction t the minors and nolr'
 '1

.^*
g.

la
lJ'u
,:.

u*fut mrrsmiiilon tfrougtrtne mirrors.'Oicillations opcur inthe lasercavityover

I
of frequencies where the cavity gain is sufficient to overcome the
a small rauge
:-

abovg lqssgs,.Hence thc device is,not a perfectly monochromatic sgurce brrt

r emits over a namow spechal band. ThE central frequeacy of this spectral band is ' . '

determined by th. ** *ogg, level difference oftbe stimulated mission trartsition.

.Other oscillation frequencies within the spectal bands result from frequency
variations due to the thermal motion of atoms within the ampliSing medium and
by atomic collisions is knowri as Doppler broadening. Doppler broadening is
'=

refgrred to as an in homogenbous broadening mechanism since individual groups

of atoms in the collection have different apparent resonance frequencies.
The optical radiation within the resonance cavity of a laser diode sets up a
pattern of electic and magneticfield lines called the modes of the cavity. The
modescan be separatedinto trro independont sets of traverse electric (TE) and "

tansverse rnagnetic (Ttvf) modes. Each modes can be described in terms of

'longiturliiral,
taterategd tansverse talf . sinusoidalvariations ofEM-frelds alogn* .
the major axes of the i*itv. '
. Longitudfncl modes: Ttrese modes ane rel.ated to the length "t" of the cavtty
and it determines the principal structure of the frequency spectum ofthp emitrcd
' optical.radia.tion. Since the length of cavity (Z) is much larger than the lasing :

I Leteral modes: These modes lie in the plane of the pn junction They depend on ,. r'

'
tie side of wall preparatioh and the uridth of the cavity alrd detennine the shape
,. of the lateral profile of the laser beam.
Tralsverse Bodes:These are associatedwittr
mqdes the eleetomagnetic field
and beam profile in the directioh perpendicular to the plane of the pn junction.
Theydeterminalaselcharactoistics suchas the ra$ationpattern and the threshold
current d6nsitri.e., thq point at which lasing starts.
Ilser Diode Rate Equations:
' The rate gquations govem the interaction of photons and electrons in the
activercgion are used to examinethe relationshipbetweenopticaf output andthe
' diode current for a p-n junction with a carrier confinement region of depth 'dt.
n'q'i are given as
Ihe rate equations'for electgn density 'z', and pheton density
dn J
.7 E.i.t.-
i

dt qd'*'"r ... (2)

 s.
ree'Examination

... (3)

C is the co.efficient describing the strength of the optieal absorption and

intelltionl t *.is spontancous recombination li&time, t is the photon
! :.Tyon *
lifetime and '7' is the irijectitin - current density- The first term iii equation (2)
I represents the increase in the electron concenkation in the conduction band as
f
I cunent flows into the device. The second and third term gives number of electron
J loss from the colduction band due to spontaneous and stimulated tansitions.
1 similarly the first term in equation (3).is the source of photons resulting from
I stimulated emissibn. The second term represent the number of photons due to
] spontaneous emission andthird term corresponds to photon lpss.
- m.ttof'the
[ -behaviour 1a!e
eeuations may be used tg study both transient and steady state
semiconductorlaser. But steady state solutions are very irnportant.
]
fr.
[F zero.stgady state solutions are foq4d by equating
lift hand side of bquations to
For any value of q , de ldt wifi only be positive

t
co'G >0 ... (4)

Equality of equation (4) in satisfi ed by a threshold value of tr t. if " n" is larger

,funthisthrcsholdvaluo, therr"g"caninorease,thethr€sholdvalue fortheeleiin

1
' , nthq (ru')' ... (5)
Tph

The thresholdvalue can betlkpressed in terms ofthe tbreshold current density

!. to maintain an inversion levdl fl = nth in the steady,state ivhen ffi
number of
I =0.

ts'-=b
fil,
qd
I
The equation (6) giveq the current requited'to sustriin an excess electon
ity in the laser when spontaneous emission is the only decay mechanism.
In the steady state condition at the lasingthreshold, the bquation (2) and (3)
be udtten as,
 tical Comm.uniqalion

qd 7,p
:

' i t"(8)
: . :. ','l

solving $e equation (?) and (8) using (6) gives ' I , ' which is equal to ,

..'.'(9)

' -r ' resuttingfrom i 'ge*crated '' ' . :

dtir.nutateaemissiolt Photonst

13.o)(l)Whrtt}?eofmaterielsareusedforopdcalsources.Whatare
-
ii;';avrntiges of double llglerl stScturl,.Qomptre surface
: emltting rnd edge emlttlng LED structures' (8) ' :

.for
Ans: Maior Requirements an optlcal qource (or) Optical Fiber
Emitter. ,.

,(l) |islro.tnlshyldbelishtrryIr*,,
.: \
nimizedistortion
: :
(ti) - SourcessUorlai'bito"'thceiecuicalinputaccutqtelytbry, :

(lD Sourc,es should emit light atwavelqngths where the.fiber has low losses

(tg So,*o *rrfadirectlymodulatefte signal voerawideban&ridthextending

' (r) Sotrcesh$dhayolanrqwspggealwidthorlineu'idthominimifpdisper.sion

(vD Optical output must be stable irreqpective of clranges in ambient conditions

' (vll) Sorlrce size and Confi$Iqtion

;,
mustbe cornpatible with ur optical libet for

.tdvanage of double Heterostructure:

, ftro differeiii alloy laye-rs'on'each side of $e active region is confined in
gonln!
a.rrrUfei"tero rtn tt *.'fUtband gap ene,tgy differenqe ofadjaeentlayers
' tHe csrriers and re-ftactive indices difference of adjoining layers confine
"narge
 ]?ES siEI

.
*
,lr

?:':r:
_
.'
t
A0 ' .B,E,n.Tech. Desree Exaai :

rUe opieat gted to&Jcenral laycr. So, higlr qfficiency and higlr radiance obtgrned
is
ductothi.qdual*onfinemeRt.::..]-.:
poiuparison between surfaee emlttlng rnd edge emlttlng LEII
l.
I
The life time of cdge emitting LED isabottt 105 Hours.

[, The edgr emitters have lower &ive current'than the.surface eilritters. ,

l. ThemanufactringofedgeemittingLED ismoreexpensivethanthesurface

l. For edge emittlrg LEDs the spectal width is more such that it is 50 nm at
i O.ss [rmwavelength snd 70 nm it t.g pm wavelength.
13. o) (it) Derlve rn expresslon for the lnternrl optlcal power level
,..,.generrtedtnLEDs!.(8)':
;t
.. . t is the total
R- number of photons generated per second and each photon has

" ' : , hcl ,.

E P. = lin,f I qhu = Irr -
FntqL
! --

Thepoweremitted from LED isnotonlydue to internal quantum efficiency

E
f"t itir alsb,to external quanhrtn effrcienpy tl*'. It is defined as the ratiorof
poton emitted from LED to the number of internally generated photons. All the
Emnally gerl€mted photons will not exit the device' Some$mes the external'
@ntum elficiency is dcfined as thc.ratio ofthc trumber ofphotons ernitted to the
5tat numb,er cf c,arrier rcpombiriation.:Extemal quantum efficiency cirn be found
'fr,om the figure 1.
E
-
"*i C'en flnc rncht' l.q)ror
F
t:

Li!&f*icnrntlnl
H rad
'
rclt*n'
::-
r-
ryidl.ng
-:

tctl*ttd wtYa
(itnitntrrrrint lrttrt
t-
i-
:. ' Flg.l .

The reftaetion effec.ts at,the Surface:ofthe LED are'taken into acoount. At

 tical QqSmunicatiqn (EC-l qOZ)

g" =Jr lZ-g, =sin-'t(n, lnr)

.'
Where nt' Refractive index of the semiconductolmaterial
-ov

.Qc
,fho xternutclu4otumefficiencl' e*r.=f Jtt+Xrrsin{)dS '
... (3)'

' I( tp ) ' Fresnel tansmissioa co-efficient or fresnel tansmissivity. This depends l

4flrfr2
Eor normal rncidence r-{ g ) =' ?I0) =
fu
nr^\ 4n
(o)= (,,.)1
Substituting equation (5)in equarion (3) ,: ,

Then'Io, can be written as

,a

u
\,*,=AT.*, (rnsin s) ds

=6'ffif[-cos0]f
I

2n
*Gf[:cos(0')+ cosoJ

2n
= tz'q,r.
ffill-cos(n '

l*tr
n{n+ l),
The optical poweremitted from the LED is

P=QrrrPin, = :;ii'm
 rlo,hcl
, q\"n(n+
i\u
1)2 ?

14. (a) (i) Exptain the operation of avilanchephotodiode. (10)

lf. (a) (ii) pirtlc,tlar dliclU

The quenru4 ef,i$ency of a ir 80% for
IAPD
: the detection of redirtion at a wavelength of 0.9 P m, when the
$ inciOentopticalpoweris0.S pW.Theoutputcurrenf from thedevlee
i (after avelanche gein) is 11 pA. Determlne the multiplicetion factor
E of the photodiode under these condltions. (6)
lr]
ffi Ans: Multiplilation factot =M - -*
I,

rr=RPo:
#.Po=ff.n
=

[=0.289,FrA i .

1lF{
M_ 0.2891L{

ld = 38.06.

O) (i)
Drhw thecircuit diaglam of htgh tmped$lce pr€'8mplifier and
explain its operation. (8)
Ans: High impedairce preamplifiers are used to reduce all the sources of
to-the absolute minimum. Low-input capacitanceo detector with low dark
are used to reduce the noise. The thermal noisc can be reduced by using

-impedance
aqplifier. The high imped*t. Ptogytt: ulT,g1npqt Rc time
nt and the front endbandwidth is less than a signal bandwidth. To compensate
ir ti lnput signal is integrated and equalization techniques must be employed.
b.high rmpedance preamplifiers canbe furtlier classified into.
 tical Communisatio& G&t193
ri s.
T,
i) High impedance FET amplifiers and

tt)'High impedance bipolat hansistor amplifi ers.

High impedance amplifier has the following fqatures:

:(tuLargestsensitivityl....-,':

.. \

Fig. I Simpte hlgh impedance preamplifier deslgn using FEt

FET has,very large input rcsistance "ftu" usually greater than 106p. FoI
,

Then the total resistance given bY

111 3i. (u
R&ftr,
-l---

tr,

E:& *f"' &= oJ

The principal noise sources are:

Itt Thcrmal noise fromttre lgadorfeedtackresistor,

I, ..

:" (A
 'Jrls
q.
B.E./B.Tech.
= \algQrc
.I; isthe gate leakage cunetrt ofthe FET, thethermal noiseofthesonduction
ctrgtuiff resistance.is Ch.tu.t tir.d by'the Aans conductance &.'Ite voltage,
noise spectal density is
' u
s-=*S,
_E g^
,, ...(3)

Where '?' is the channel-noise facJor. It is a numerical constant. The therrral

.:
n, - hlru,*,,ff .ffi]r lT)' Wr,B . (4)

isminimiredbyhigh impedance amplifier

The noise in ahigh impedarrce design
design which is known as integntion-differentiate apProach. In this metho4 the
detected otrfut sig4al is integfated by arrrplifier input tesiisAnce arid to compensate
this the siglral is differentatedinthe equatingfilter;Theintegrtor, Afferentiator

i
Otfiartntiator
q rr * E I
- ---
I
-tl
Cr r

$'rtnp*unsp

Qvcr rll ryste with amplifler

l'rr.etluency

I
-t*r

t4, (b) (ti) 'Discuss the source of errors in the optical receivers. (8)
 &

Ootical Communication {EC- f 402 }

Ans:

fihro$o
$rrk

* QuantumNotiie * Bulkdarkcunght ,* ThermalNoise * Amplifier

Noise

Errors in the detection mechanism arise from various sources in the signal .

, {etqg,Soo s}rytem. 'Noise" means the unwanted components of an electricaf signal '
aat taia to disturb the Sansnrission,and processing of the signal. Noises cannot., ..
| ' .ij :'
te contolled sincb some of them are random in natue. Figure sho-w,s the t
nrttry
various eror sources.
;

'Ttere are external noises which are developed efiernal to the syston and
intcrval nojses wiich [r-e developed interhat tb the system. Let us consider only , i
r tbe ilrternal noises. From the fiber optic liurer the detector receives the pulses with :

symbofls '1' and''Oi. Each 'l' will contain vari{le ntimber of of pholons. That it '
self develop quanturn noise (or) shot noise w.hose fluctuation is given by poisson
distibution and is random in nature. In the deteotion system, we have variety : of
noise like bulk dark current noises w-hich arise from the surrounding light and .
surface leakage curren tnoise from the improper zurface finishing and uniformity
ofthe surface of,the detector. Fuither there is a gain fluctuation noise in the case
of avalanche photodiode; Thlnmal rioises are arised from the load resistor and .
amplifier components. The thermal noises are of a gaussian nature and one can
determine the developed error easily. Rarrdom arrival of photons at the detector ,
gives quanfum noise or shot noise. The actual number of electron-hole pairs 'n' I '

tbat are generated

lllal a[9 B,gllgf 4I9(l flrretuates from the
lrUf,,l,Ual]ED llulll alvgr4Eg ol&\llurrrE
lrtF average according to ul9 poisson
rY the distibution.
lralDDvll urnurusuvu. I
The random nature of the avalanche multiplication process gies rise to anothedl
t,

,(

For on APD with gain 'M' and ionisation rate ratio 'k', the excess noise .

frctor F(m) for electrons injection is given by,

 .F
g.
26 B.E.rB.Tech. Deqree Examination (Apr.A{av,aOO8) I

,. F(M)*= KM +(, #) (l
where x ranges from 0 to 1.0 depending on the photodiode material.r.
Dde to dispersion or pulse broadening in the optical fiber, there is'anothel
enor souroe called 'intersymbol interference. Due to pulse spreading, some ot
, the fransmitted energy of a given pulse will progressively ppread into neighbouring
pulse resulting interference of the signals.
15. (e) (l) Exptain the ltnk power budget ln optlcal communlcation system.

, Ans: @efer NovemberlDecember 2OO7).

'f5. (e) (ii) Explatn the prlnclpte of WDM. (8)
Ans: An interesting and powerful aspect of an optical communication lin'.. ir
, that many different wavelengths can be sent along a fiber simultaneously in the
-'
l30o - 1600 nm spectum. This is achiev6d tbrouglr wDM (wavelength divisior .
multiplexing). combining a numberof wavelengths on a same fiber is known as
wDM. '\P'independent optically formitted information streams each tansmittec
atadifferent wavelengthare combinedwith optical multiplexer and sent overthe
) samefiber.

i\ l),,u
tirt itt

llrt
l**mrrw**J

i
Frg. I A unldlrectlonalYUDM ryrtem that comblner N Independent lnput slgnals for
'
F transmisElon over a single fiber
p:
Eachof the sEeams could bg at a different dta rate. Each information strean
F
lnaintains its individual data rate after being multiplexed with other sheams ant
ates at its unique wavelength.
The basis of WDM is to use multiple sources operating at slightly different
engths to transmit several independent informatioh streams over the same
 tical Co**uni."tio* (EC:lffi
ft"p".i, rhe application of wDM is to upgrad"H#::[*i,l
iPstlg:,
to point nu.r optic hansmission links. Each channel
su1 .
oirting poirt

iU Transpa4ency: Without need for a common signal t*t5;f,;Hi:l

*ur.trrgtht *"o.rd to said fist or slowasyncbronousrand syneh
f.i"-ri *afog informations: Each optioal channel can car.ry any transmission
flormat. :r .
wavelenglh '
iii) Wavelcngth routing: Wavelength routedhetrn'orks rlsethl ltual
f"rt*t ;fr. int rn iiut" ornnJaaAess. Wavelength Tnsitivqoptical routing -
i.vrri *; oJ in wanetengttr routing. Wavelength is usedas another'dimerision'

Operatlonal PrinciPle of WDM

-'-*"0.,
f..tur. ofWpfvf is that the d:screte Talefn$. !s f9y t"$g"i:t
tetofcarriers that canbe separated, routedandswitchedwitholtinterfti'qY1ft
characteristics of a high'
:ach other, To have clear idea of WDM, the spectral

Thecurvein.figurg2showstwolgwlossregionsofasinglemo!!|fer
..xteidover the waielengthq ranging frbm 1270 to l350nm and from 1480 to
,;bil. mar. regions r* u. viewed either in tiiruts of spe'ctr-al width or by .

roeans of optical bandwidth-. :

deviation
The optical Uana,,riatfr can be et'pressed in terms of wavelengdr
"a?L" :

r^r,=[#) ro^r ,., ( 1)

Since the spectal width of a high quality so-urce

occupies.only a narrow
,riJu*o*i,n. ru. *o low loss windows provide many additional operating
t'
 -EFrT:roa.
t
qp
^rs

.Tech. De n*l*ination A 2008

laser is locked means the frequency of the laser is fixed.

lnpul
.:!x**tl Sinslc liber
L-.
*.
l.
E
&rtpur
r:hrn*tl

Fig. 3 Schematic rcprcsentation- of a bidlrectlonalWPM systom in whlch two or more

r wavclengths aretransmitted qimultaneously in opposlte directions overthe same fiber

t A variety of active and passive devices are reguired t9 implement WDM

I netvriorks. Passive devices require n6 external control for their bperation..The

f
active,Ur^r*:Ti be controlled electonically.hence rpovide a lrge de8ree of ,

F At the hansmitting
end, there are several modulated light sources. The i :

p muhiplexer comtines these optical outputs into a serial spectum ofclosely spaced .

I wavelength signals. At the receiving end demultiplexer.is used to separate the

, optical signals into corresponding channels for signal processing. Photodetectors
are sensitive over a broad range ofwavelengths. To prevent spurious signals the
demultiplexermustexhibitnarow spectal opemtion orvery stable optieal filters
with sharp wavelength qut ofPs mulst be used.
O) (i) Explain the basic.,lconcept of SONET-netowrk
15.
Ans: (Refer Novembei/December 2OO7).

l
rD

I"
 - " - - - -.=;r+-,,*@f,!F..
".
. ,,6.

r#

C 326;4

Sowenth Serneste

Electronics and communication Engineering

EC1402-oPTICALCoMMUI{ICATIoN
!^- tnnE\

Answer ALL question$'

PARTA-(l0 x 2=20marls)
the ray optics
1. what are the advnntages and disad.vantages of
theorY?
and refraction
Ans: The i*g. scale,optical effects such as reflection
process of ray tracin$'
. . - F..^-- A..r^Ica

crin be analyzed by simple geometrical

me light beam'
The rays show the direction of energy tlow
flow ln
in the

The wavefronts are separated by one *1]:tength'

^
for an optical
. 2. A typical relative refractive index difference
is 1%' Estimate
fiber designed for long distance transmission
the core index is
the numerical apert,ri* for th(r fiber when
1,470
Ans: Given: Core index -- l '47
'.'. t}, = l'47' - ' ^r
I
lt r r
.

nl-[?=0.0 I ['.'tr,-o, l%]

r.'. 1.47- 0.01 = n*
[1. = 1.46
r ()' 17 t I
NA= -nfr 6*ilz '(1.46)s
3. Define gnouP delaY.
each speotral component
Ans: $s the signat propagited alols *1. fiber
.can be adsumed to travel independentty and to underso a time clelay or
 Yffi'.'..%n!E
..9
s.
2, ,
t.

B.E./B.Tech-Deoree Examination (Nov/Dec 2008)

-
group delay per unit length in the direction ofpropagation given by
Jr& = lA/g * l/c * dp/dk = -?'2l2nC * dp/d}..
4, A multimode mode graded index fiber exhibits total pulse .?"
broadening of 0.1 Fui over a distance of 15 km. Estimate the :
maximum possible bandwidth on the tink essuming"*.Z
coding without intersymbol interference. .f
Ans: Given:
o = 0. I Fts !

d= 15 km
. o =(ointerrooatl+oinhmodar2)l/2.
5. Distinguish direct and indirect bandgrap materials.
Ans: [n direct bandgap materials, the electron and hole have the same
momentum value.
In indirect bandgap rnaterial, the conduction band minimum and the
valence band maximum energy level occur at different values of
rnomentum.
6. An tED has radiative and ilon-radiative reeombination times
of 30 and 100 ns respectively. Determine the internal
quantum efficiency.
Ans: Givent r = J0 ns, r r, = 100 ns
T,to,. 80 x 100
; = ----=-
xr+rnr= 80+100 --t-
= ?3.1- nS

r -iiS
28.1ns
r in, = r, = so1,q = oJ7

Ans: It'is defined as the ratio of output photo current to the incident
optical power"
ft = Irffo- nqftrv
where ft = Responsivity
Ip = Output Photo current
Po = Incident Optical power.
 .,B+
g,

OpticalCommunication'(E0-1402) 3 '
.-:
I 8. Compare the performance of APD and Pry diode.
Anst, APD PIN Diode
1) The carriermultiplication- When an incident photon
yields newly created carriers has an energy greater than

which are:also accel6rated bV I or equal to the band gap

high electric figh thus gai- energy, it give up its
energy and exc.ite an elec-
t

ningenough energy to cause

impact ionization which is tron to conduction band.
knoqn as avalanche' effect. This process generates the
photocarriers.

ElffitioFhld
+

p * Aualar$e

i""'treg'on
i(n) tfinlmum Frdd

fuquiredhr

--lm$:lo$lon
p

Analogous to the pio photodiode the performance of an APD is

Ro.o: ffi -ffiJ\r*

9. What is the significance of rise time budget?
Ans: A rise time budget analysis is a convenient method for determining
the dispersion limitation at an optical fiber link. This is particularly useful
for digital system. [R this approa.h, tr* is given as,
r1/2
/ N
t,*=
[tri]
 ..9
g,
B.E.lB,Tech ree Examination (Nov/Deq2008

10. Compare doped fiber amplifiers and conventional rePeaters.

med FiberAmplifiers C6nventional Repeaters
1) These are needed PeriodicallY Conventional repeaters are alsP

for ampli$ing and reshaPing the needed periodi cal ly for, qmPl i-

signal in [ong-distance links. ffing and reshaping the signal

in lond-distance links.

2) These hre used as front and Repeaters are used to increase.

amplifier for and as LAN booster the signals strength for a

icular distance.

.a

11. (a) (i)' Expain with a neat block diagram the fundamentals of

Ans:

F!g.:Majgrel9mentsofanopticalFiberTransmissionline
An optical fiber ffansmission link comprises the elements as shown
in the Figure. The key assumptions are transmitter consisting of a light
source and its associated drive circuitry. Additional compohents include
optiOal amplifiers, connectors, splices, couplerb and rygenerators.
Analogous to cgpper cables, optical fibre cables can be installed
either aerially, in ducts, unde$ea or burie"d directly in the ground. For
undersea installations. the snlicins and reoeater-installation functions are
 - '-?{'i6='*t-
g.
:
'.F
g.

Ootical Communication tEC-1 402) 5*

; camied out on board a specially designed sahle-laying ship.
t Early technology mode oxclusive use of the 800 to 900 nm
, wavelength band since, in this region, the fibers made at ttrat time exhibited

' a local minimum in the attenuation curve, and optical sources and I

photodetectors operating at these wavelengths were available. This r€gton

is referred to as ihe first window, By reducing the eoncentration of
, hydroxyl ions.Eld metallic impurities in the fibre material, they were
able to fabricaffitical fibers with very loss in the ll00+o-l600nm
region. Two windows are defined here: the second window, centred
around 13l0nm, and the third window, centered around l550nm.
" , Once the cbale is installed, a light source that is dimensionally
compatible with the fibre oore is used to launch optical power into the
fiber. Semiconduitor LED's and Laser diodes are suitable for this
purpose, since their lighl outplt can be modulatAd rapidly by simply,varying
the bias current at the desired transmission rate, thereby producing an
optical.signal. The electric input signals to the transmitte circuitry for
the optical source can be of either analog or digital form. For high rate
'systems, direct modulationofthe sotrce can lead to unacceptablesignal
, distortion. In this case, an extemal modulator is used to vary the amplitude :
of a continuous light output from a Laser diode source.
.After an optical signal is launched into a fiber, it will become
' progressively attenuated and distorted with incroasing distance because
of scpttering, absorption, dispersion mechanism in the glass material. .

11. (a) (ii) Discuss the mode tneory of circular waveguides. (8)
Ans:Inthis,itisnecessarytosolvethe|r4arwell'sequationsubjectto
' the cylindrical boundary conditions at the inteiface between the core
and the cladding ofthe fihr.

only transverse electric (TE) modes and transverse magnetic (TM)

mcdes are found. This gives rise to,hybrid modes, which mekes optical , ,
waveguide analysis more complex than metallic waveguide analysis. ,.,.
. Although the theory of light propagation in optical fibers is welli. :
underslood, a complete description ofttrq guided and radiation modes is '
.
, rBthercomplex since it involv.es six component hybrid electromagnetic
fields that have very involved mathematical exprcssions. A simplification
 :-,:Qs
,.8
'e.

6 B.E,/B.Tech Deqree.Examinaiion (Nov/Dec 2008)

ofthese expressions can be carried out, in practice, since fibers arp

' usually constructed so that n, - q < < l. The field components arc
called Linearly Polarized modes are given as Li, where j and m are
integers designating mode solutiotis;
Although the ahalysis requiredforeventhese simplifications is still
fairly involved, this material is key to understanding the principles of
: optical fiber operation.
ri, 61 lry Discuss briefly abo-ut lineerly polartzed modes. (6)
Ansi In practice, since fibers are usually constucted so that trr- ha ( (
1: The field components are oalled Linearly Polarlzed modes (LP)
and are labelled as LP.;* where j qd m are integers designating mode
solutions.
In this scheme for the lowest-order modes, integ€rs of each LPo*
mode is derived from an IIE, mode and each LP* mode'comes from
TE* and TMo,, and [IEo, modes. Thus, the fundamental LPo, mode
corresponds to HE,, mode.
An important parameter connected with the cut-offcondition is the
V number defined by

The V numbers can also be used to express the number of rnodes

M in a multimode fiber when V is large. For this case, an estimate of the
total number of nodes supportod in a fiber is

Mn,
LW).,"t-ni)+.
At the cut-offpoint, the mode becomes radiative with all the optical
power of the mode residing in the cladding
P4
r ccad-
^- t

Note that since M is proportional to V2, the power flow in the cladding
decreases as V increases.
 ...,..:=s-g::..:.......l.:..-..--*:::':'1.=:-.ffi€-jaelj::-.-:*,".

gdiggl.g.o,ltl,rBgl[tic3tign (EC-1402] . .- ...-- ; -. ,

:- 7
i

Il. -
(b) (ii) Draw*,the structures of single and multimode step
intlex fibers graded index fiber with theil typical
"nA
(6)
dimensions.
Ans: Typical Dimensions
Fiber Cross Section & RaY Paths
siI i:' I
(Cldding)
125 pm

i2a
ti-t- -T- &12um
lCorey

f*T T
J Monornode Step lndex Fihr I 12s4oo pm

II
(Cldding)

--r-
.- *
50-200rrm
(Core)

Multimode Step-lndex Fiber

I
I

125'140 pm

tI
aaddins

I so-1ooLm
-T: cladding
MultirMe Graded-lndex Fiber
-.- ----^l^.t t !-
Fig. : Structures of single and multimode step index fibers and graded index
fiber with their $pioal dimensions
(iiD Mention the advantages of optical Iiber communication
11. (b)
i systems. (4)
I
&
B
Ans: l) Attenuation makes wide range of distance possible
T
t:
r
r
r
;

4) No physical eldctrical connection is required

5) More reliable

7\ Greater bandwidth

9) Data rate is much higher

10) The cost per channrt l, lower
 ,l7t;;tr*{]@6.

*
."*

B. E.lB.Tech Deqree Examinatio@

ll) There is no radiation and interference on other circuits or*

i.

!-

l,

F
ts
:
I'

h.

F
 (opz)z = .
[H)',ur, ;[H)sGi
Thus the scattering losses vary as X-4

#
tt
-
.E

E
ll---r-r+!
Wavplength
(rrm)

Bending Losses: Microbending and macrobending are two types of

bending losses.
Microienaing Losses: Microbends have small random deviations
about a nominat straight line position. These are observed in cabled
fibre. This cross-coupling leads to the loss through cladding. On some
occasions these bendings may be quite sharp and also will be introducing
ceftain loss of light energy in the mibre.

For single mode,

n n p Kn,(Fd)6(NA)4
onic = u.u) clm

where, 0* = attenuation constant

t6 = wave vector = 2tl7t
a = core radius
Fn = half mode field diameter in the single mode fiber.
For multimode, \\

cmic=Qr[#] .[*]
 t

10 B,E.lB,Tech Deoree Examination (Nov/Dec 2008)

---
F
aE

E
E
-
.9
l-

-E
"fu E

5
g
o)

-r--1,
Wavelength
t(pum)
/

Mabrobending Losse$ The larger the fiber core radius and smaller
the bend radius, the'greater the macrobend loss.
' f ol.r\1

0*.. E - I o rog
[' #{+. [#)"']]*' f

. where P = lndex profile parameter

R = Radius of curvature of bend
:

K = Wave Vector = ZnlL

( , za .( I )"t'] I
:

[,
cr---=-lologl'-l -[.2**fl] %R
mac
L \-- -- J.rl
$
.s
I!

€
E
I
Bandlng Dlameter ln mm

12. (a) (ii) Discuss polarization mode dispersion and its

-r-:

,
limitations. (8)
,rarll\rrr\rl \v

Ans: Birefringence can result from intrinsic factors such as geometric

irregularities ofthe fibre qore or internal stress on it. In addition, external
factorso suctr as bending, twisting or pinching of the fiber can also lead
to birefringence.
*rt
"
 's,

Optical Gommunication (EC-1 402) 11

i;
r
r
p
g

F.

t!.

b
p
$
il

$
!
r
k,
::1"

,s
.- 11

It
+!
._,:i-

?f*
t*"
'*11'
.*{:;
i:;'
-.*,,1.

$
;I..
"ffi
s&iF
q'
i**

x&
ffi
dFsS
ddffi'"
f"l''
 : r=his'jj.WEfin

*
. r"rt

\$-

.r 2nn(1,)
P= r
L( dn)
.'. tps; = - t\^ dS
Ct, J

d."t"lkl"^
.:
=*ii-l
=e^L[^el
c L'- u^, 1

= crl, I D*"t(I) |

where
' D**(I)
maf. '
is the material dispersion.
Waveguide Dispersionl The effect of waveguide dispersion on pulse
spreading can be approximated by assuming that the refractive index of
the material is independent of wavelength. To maks the results
independentoffiberconfigurationconstantbisdefinedas

gz lKz ,- n!
$= *i '.- *i
For small values of the index difference A - {n, -nr/n, it can be'

B/K"*n"
bns6f; Bnr nrk(bA+ 1)
it-

t;

Group delay t*, arishg from waveguide dispersion is

Ldg L[
ro*::*==r
t
Cax-clII2TII2A^d(Kb)l
dk I

Which is valid smi values off A.

dlid for small A. '

LT
r*s =
c 1",
+nrA#]
 .-.€F".-.--{-WE
-.F
s.

Ontical Communication (EC-1 402) 13

The factor d(VJdv can be expressed as

' doro) _ rf, _J,*1(uaF,_r("")l

z;3(ua) I
- -L-
:

; dv ,.
1

" Where u is defrned and a is the fiber radius. For multimode fibers '

the waveguide dispersion is generally very small compared with material .

dispersion and can therefore be neglected.

12. (b) (ii) Write a brief note on pulse broadening in graded
index fiberc. (15)
Ansr The analysis of pulse broadening in graded:index waveguides is
morc involved. owing to the radial variation in core refractive index, The
foatunc of this grading of the refractive index profile is that it offers
multimode propagation in a relatively large core together with the
possibility of a very low intermodal delay distortion. This combination '
allows the transmission ofhigh data rates over long distances while still
maintaining a resonable degree of light launching and coupling ease.
Since the index of refraction is lorrer at the outer edges ofthg oore, light
rays will travel fcstertothis region than inthe i" of the core where
""n
the refractive index is higher. This can be seen from the fundamental
relationship. V = c/n, where V is the speed of light in a medium of
refractive index n. Thus, the ray congruence categorizing the higher'
order mode will tend to travel further than the fundamental ray
congruence, but at a faster rate.
The root-mean-square pulse broadening <$E sigmD in a graded
index fiber can be obtained as
o = (o'"n**, + olinorrnod"r)l/2

o,nren,,odar = (r.31 -(..)')"'

L o?,,, ,n g d2[r
6 '.2
inremodal
= eT(-(A 1;E--)- -

- NrcrAlzl,'d'*r o -Nrcrax[.o*zfu}"
 _ 3_rh;E-.:..
E
.f{
s.
. .1t ,- ,.., , ' , r.
, FiE,ts,Rch DeorceWnation Nov/Doc200g) _.-*
, .13.'(a)'(i) Draw and explatn'the differeri rt*ttr"". used to
; achieye carriel and optlcel contrinement,il,leseiAiodeu. (8)
i Ans: For optical fiber communication systems rcquiring$andwidths
: greaterthan ipproximately 200MHa the semiconductor injection laser
diode is prefened over the LED. Laser diodes typically llave rcsponse
time less than I ns,have optical bandwidths of2nm or lesq an4 ii generirl,
are,oapgble ofcoupling several ten ofmilliwatts of usefrrl luminescent
i- power into optical fibers with ryall core and small mode-field diamerers.
Virtually all laser diodes in use are multilayered heterojunction diodes.
i . The'dou,blo he.torojunction LEDlconfiguration evotved,from the
successful demonstntion'of both cerrier aia opUcal .oono.r.it ii
r heterojun
Qn rlirjection iaser diodes. The more rapid evolvement and
utilizationof LEDs as cornpared with laserdiodes lies inthe inherently
simpler construction, the smaller tempenaturc dependence of tre emitted
optical power, and the absence of catastrophic degradation in LEDs.
I In other laser diode type, comr-nonly referred to as the distribuied-
L feedback laser the cleaved facit are not required for'optical
feedback.

Ir
i:

l-

F
;; FiS. : StruUture of a DFB laser dlode
:
E_, The stimulated emission rate into a given mode is proportional to the
. intensity of the radiation in that mode. The radiation intensity at a
L

f,,-
photon energy hy varies exponentially with the distance z that it
traverses along the tasing oavity according to the relationship.

F
where il is the effective absorption coeflicient of the material in the
r".
:
optical path and f is the optical-confinemetrt factor - that is, the
I fraction-of optical power in t-he active layer.
k-
 i-t<F*[_ *iBEE*n.

, ..ft
g.

OptlcalGdmmunicationlEC-1402) . - 15

of temperetune op the-performlnce
),7'

ffieeffects
'of a leser diode. (4)
Ans: An irlportant factor to consider in the apptication of lasor diodes
is'the tempeiut * dependerice ofthe threshol! culent This
parameter
because
inffeases with t6mperatue in al! types of semiconductor lasers
, of various complex temperature-dependent factorq, However,.th9
ternperature variation of.l* can be,approximated by lhe empirioal
expression.
Io(T) = I, Gt^
, lilhereToisameasureoftherqlatitetemp"dqfu'TTsitivityand
di{et
I- is a constani. For a conventional sfiipe-geometry GaAlAs Laser
q jf" An
is typically 120 l65qC in the v!9lniU o{1oo*-luPperafure'
-
u**p' of a laser diode with To = 135"C and I" = 52A..lhe:t :lt:t
' dupgid.r"es of I* on tempepturi have beendemonstrated for laAlAs

Fsr these laserg Tu can be as hig! as 437'C' C6nsequeryly; if a

constant optical outp, pt pgw-er: level is to be rnaintainea as tht
*:.Tptt$*
it is neoessary adjustthe dc
of itre laser.changei o.ur the laser dges, 1o

bias ourrent level.

13. (a) (iii) Glvea brief aecount of the modulation of
en LED. (4)

Ahs: The frequency respoise of an LED is largely determined by the

following three factors; the doping level in the active region, and the
drive current is modulated at a
nar.asrtic-capacitance ofthg LED.If.the
it .;u*.! o;.the power of the delice will vary''
. "p,i"*ur-put
P(co) = Pu [ 1 + 1ol*)'] -ttz '
This delay is negligible if a small, consta nt forw*r; bias is applied
to

the diode.
Normally, electrical terms are used since the bandwidth is lctull{
determined ,lur the associated electrical circuitry. This is the
elechical
i-ag point; that rp, the frequency atwhich the output electrical power is
reducld by 3 dB with respect to the input electrical poy.r.
i .
electrical power
Thus, since P(or) = I2(ro)lR, the ratio ofJ\e lutP:!

F
 -.-eg_..,',%I!E
's,
16

I
f p(ro) Il'(*)l
Ratio.,.. = to log = l0logLmJ .
Li,idl
Where I(o) is the electrical burrent in the detection circuitry, The
electrical 3-dB point occurs at that frequency point wherp the detected
electrical power P(ot) at P(0)12.
7

=r=o.To.=ffi
ffi
Since the detected outrent is directly proportional to the optical

' power, this ratio is Ratioootic* 10 log

lprcoll
- r0 log
I tf*ll
.
/' .' 1
L(0) I Lr(r)l
13. (b) (i) Derive expressions for the power coupled from a
surface emitting LED into step index and graded index

Ans:Thecoupledpowercanbefoundusingtherelati,onship.

p= Jae, Jao,n(A,,Q.)
A, .Qr

r* 2n 2rE 0o,,un*

00 0 0 'i

Where the are and solid acceptafge 6ngle of the fiber define the limits
I.
ofthe integrals. The total coupled power is then determined by summinq
up the contributions from each individual emitti:ng-point source of
t
:"
F'

incremental area {Q r dr; that is, integrating over the emifting area.
!
.nf

r. 2n on, qu*

p=
J Jtznno
i00
I
0
eos O:sin 0de)de*rdr

t" Zn ..

= FBo I Iua24o.rdr PLED.step = Tczf eo(NA)z

00

= Znzrfnoniza
a
 '.
s.
"r+

17
s
a

of aeaA, into a hemisphere (2,n Sr); This is given by

.\ .t--ro
, .. 2xnl2

' *l

t steP t s\'\r L/'/

r'
LED,

P,,o'st€p=[*),n.*A)2ro,r,}a
ri

index fiber becolnes (for r, ( a)

P

--l'\
n! P ot
LED, graded
= ZzczBo ltn'(r) -
0

r ., a.]

=2P,nfal
=arrs,,lal "_a;E[-;J
'-
4[s)"1l.
If the refractive index n of this medium is difference from n,
.!,

t- -12
\trr*n) i
Where R is the Fresnel reflectign or the reflectivity at the fiber:

'!

los
'-e f l-*l : - lo log (t .- R)
I )
P.*o ra '

: l0 log (0.826)
., - 0.83 dB.

' This number can be reduced by having an-index-matching material

t
 -.g,,lgra:%E
r
"l(
B.

18 , . , , "- B,E./B,Tegh Deqree E[aminatipF (,illov/D.ec 2099]

13. (b) (ii) Exp-lain the mechar:ical misalignm'ents that *"n ;il

{s l-- :

).lt
i-T ;-]
(

(a) lateral (axial) (b) Longitudinal (c) Angular (end separation)

'a

owiqg to their micigscopic size; A standaid multimode graded-index

fibpr core is 50-l00Frm in diarneter. Radiation losses result from
mechanical misalignments because the radiation oche of the emitting
fibeq does not mateh the acoeptance onc6 of the receiving fiber. The
.tY

, magnitude ofthe radiation loss deppnds on the degree of misalignments.

Longitudinal separation o.iurc when the fibers have the same
axis but have a gap S Uet ueen their end faces. Angular misalignments
results when the two axes form- afia
__;_ _- ngle
__e_ _ so that the fiber end faces are

no longer parallel. Axial displacemuot results when the axes of the

two fibers are separated by a distance d.
: 'The most cemmon rqisalignment occ.urihg' in practice,
which also
causes the greatest power loss, is axial displaeement.
14. (a) (i) Discuss with necessary expressions the different types
of noise that affect the performance of a photodetecton tiOl
Anli In fiber optic ccimmunication systems, the photodiode is generally
. required to detect very weak optical Signals. The power signal-to-noise
. ratio SN at the output of an optical receiver is defined by

S Si gnal Power from Photocurrent

N Photo det ector Noise Power + Amplifier Noise Power
r-
I Toachieveahigh'ra!iothefollowingconditionstobe.met:
L_
I 1. The photodetector must have a high quantum efficiency to
F
generate a large signal power.
r
k:
r-
I

t
t
)
 -,.:
n,.ri,.ol f'-nmmr rninatinn (EC-1402) , '-
19

i,Thm;aectorandamplifiernoisesshouldbekeptaslowas
I
The sensitivity of photodeJector in an optical fiter qommunication
a

describable detectable optical power.

mini*u* oer€:ua:
Lhe mtnlmum
in terms of the
system is
it,i'istheopticalpowerneceSSarytoproduqeaphotocurrent.
Noise Sources:
' Bioo Voltage

Thephotodiode has a small resistqrce a total.t"l"."@!L!l \, .

and a bialSsistor
consistin.i of junclion and. paokagurg capacrtan:es'
RL.
-
: If a modulated-sigpal of optical power P(t) fatls on the detectoL the

no
. --.,
i*(t) =:=P(t).

For pin photodiodes the meanlsquare si$nal ctinent

is'
(is2) =osrpin2 = (ir'z(O)

in a bandyidth
The quantum noise current has a mean-scluare value
of the photocurrent Ip:
B yhich is proportional to the average value
-(ioi)=oq'=kqbBM'F(y)'..
'. .
The mean-square value of the surface dark culrent il given by
,

^ ^ r r 2ar.t/l. i\r\

The surface dark cuffent is also referred.to as a surfacs

leakage

current.
 1ffir=.
: 8
(.r{
B

20 B.E./B.Tech

Tftthrtodetector load resistor contributes a mean-square thermal

current is

' nl"
t

(ir') =g r' = RL ''

Where K, is Bpltzmann's constant T is the absolute temperature.
14. (a) (ii) When 3 x l0rr photons with a wavelength of 9'-8i
each
'
p m- re incident on. a photodiod.: on average 1.2 x 101r
electrons are collected at the terminals of the d1vi,ce.
D.*rmine the quantum efficiency and responsivity of the

Ans: t

14. (b) (r) Draw and ex41ih the high-jipedance preamplifier

dqsigns based on B.$ and FryT. (8)

Theinputresistanceofabrpolartransistorisgiven.by,.

low-noise design, R, and \

ane chosen to be much greater than Rr, so that \ = \n:
/;
The spectral density ofthe input noise current source results from
the shot noise of the base current:
alt - 7n

'zkBT
I ...
-E
Sl*. l-
gm
,'

From'this we can get o as'

F. .,
I
I
 .-@I'QSf
'.^ld

qF

@=H[*tre+HRin,,]

eorpttrrg
Capacihr

FET Amplifierr: At lower frequencies, silicon MOSFETs or.JFETs

are $enorally used. The piincipal noise sources are thermal noise
.soEirt*a *itn the fET c[r*o*i conduetance, thennal noise tory the
load or feed.back rcsistor, and shot noise arising frory gate:leakage
cufrent. Since the amplifier input resistance is very large, the input current

The tfttttul density is given 8s,

Where the FET channel-noise factor f is a numerical constant that

accounts fsr thermal noise.

-[
.(zuc)'*KeTr
LJ s-
W; -.:=4ig *t-
f.
'.F
g.

F =
i.-
I
-

14. (b) (ii) 'Write a brief note on transimpedance amplifien (8)

Ans: Although a high-impedance desigll pfoduc€s the lowesl;noise
. amplifier, ithas two limitations d)forbroadband applicationq equalization
is rlquired; and b) it has a limited dynamic r?nge:An a-ltlnativ€ design
is tdtransimpedance amplifier. This desigp yields both low noise'and a
)

large dynamic range,

For the transimpedance amplifier the thermal noise characteristic
\Mti at the equalizer output is therefore simply found by replacing \,
RuRr
Rl =R6+Rg

w,z =
3[r,
q-\
*ff. &)', . ffiSsrs
Ki

.-.1=-*:_=::l-*-*-
--t-r

In pfactice, Ihe feedback resistance & it much greater than'tlie

amplifier input residtance R.. Consequently, Rt = R, so that.
T. 4k,T
wz= w*, * ;7 [2
qRf ":

where W*,
tS'--
i, the hlgh impedance amplifier noise characteristic, given
L
by either FET or BJT
R!-
designs.
| .

.l

:
t,
 'i"B
ry.

nication (EC-1

. Comparing the two feedback design, the transfer function of the

non-fiiedback amplifier is (in V/A).
 ,'-l-*{;*a-- a'.-
R

., ,B
g.
*
Zq' ' , ' "P.E/B.IechDedreg$xqminatlpn{N0vpec,2008l
' To derive therevolution of tle pulse shipe require! for soliton

-.jLtu=i* .N' I u l? u - j(a/ 2)u .

. For the 3 right-hand terms

1) The first term reprcsents GVp effects of the fiber.
2) The secoad non-linear term denotes the facf that the refraotive

3) The third tem. represents the effects of ene45r loss or gain.

' The solution for the fundamenthl solition is girbn by
u(a t) = sech (t) exp'(z/2).
Where sech(t) is the hypgrbolic secant frrnction. This is a bell-shaped

The time scalo is given in units normalized to the l/e width of the -
pulse.

. . d$"-r = lu(01' dz = se9hz(t) da.

F
For the notrJinear process,
'

I r azr)
d0 disn =[^F)"
\ ''

=
[;-sect"ttl] 9'
.For the dispersion effeql. Since such a phase shift changes neither
the temporal nor the spectral shape of a pulse, the soliton rdmains
completely non-dispersive ln both the terypor.1l and frequency domains.
Soliton Parameterc: The FWI{I-vI ofa pulie is defined as the full width
of the pulse at its half-maximum power level.
Thus the FWHM T, ofthe fi,rndamental soliton pulse in norrnalized
time is found from the reiationship sech2( r ) = l/2 with r:Ts(2TJ, wherc
#",
#
k

To=ffi=frn ng0.56?I
 r3..%ilEf
.^*
B
,.
a

Optical Communication (EC-1402I . 25- .v

Soliton width aiid Jpacings

I fhe return-to-zero t,

(RZ) format is used. This condition thus

constraints the achievabli bit rate. Since there,is a timit on how narrow
a soliton pulse can be generatEd. .

B=1= 1'-L'7.627
where the factor 2So:= TB/T' is the nonnalized separation between
neighbouring sol itons.
For so litons that are initially in phase and separated by 2. So
I

the soliton separation is periodic with an oscillation period:

F

\
'
-
,l erfr*O,So).
.*

Lr
I
= ffiouro
I

= Lpo.ioa exP (SJ

(zn)E c ,,A\
soll D
1

I 16 '{rt

TI
=;7ftexP(Ss;-
8SO tpzt q

ls. (b) (i) Discuss the concept of wDM with a neat diagram. (6)
Anst ",
,qt-
, 0ptlcal

3
 -.r'w,".,1i:G"
'.F
B,

To derive the evolution of the ptlse shape required for soliton

transmiSsion, one needs tg considei the NLS equation.

.Al LOzu
L+N2 lul? u- j(a l2)u,
-Je=zR
-i
i

!
V
. For the 3 right-hand terms
!\ 1) The first'terrn represents GVD effects of the fiber.
T') The second non-linear term denotes the fact that the refraetive
:
indexofthefiberdepends,onthelightintensity.]

r3)
!r
Thethirdtermrepresentstheeffectsofenerrylo.ssorgain.'.
' The solution for the fundamenthl solition is given by

. where sech(t) isthehype. rbolic secantfunctioa, This isabell-shaped

ihe time scale is given in units normalized to the l1e $vidth of the
pulse. ,. ,

(ta'u).
__r=l
dQ diso ^ ldz .

!\ \Zu -0t, )
: [r-secnrr,l] g.
L2
.For the dispersioneffec! Since such a phase shift changes neither
the temporal nor the spectral shape of a pulse, the soliton rernains
qompletely non:dispersivg in both the t qpof.t,ita frequengl lomains-
Soliton Parameters: Th9 FWIII/I of a pulie is defined as thc full width
ofthe pufse at,its halflinaximtlm power level. /' '

Thus the FWHM Ts ofthe firndamenul soliton pulse in normalized

time is found fiom tlre relationstrip sech( t ) = lD with t fT,(2\)' where
T^U is the basic normalizpd time unit..This yields

T" T"
T, =
t;,.fu' =
frdn s 0.667r,
 .,1{
a. BI,

*;
Qqtiqal0gmrnunig,a$g0jEC-J4Q?) .,; - ., -,,,-i$; .1'

.{
I The return-to-zero is used. This condition thus \.
I

6Z) format ,,

constraints the achievabli bit rate. Since there,is a limit on how narow t

a soliton pulse can be generatqd.

TI
=*exp(Se,l
8S6 tgzt q

Discuss the concept of WDM with a neat diagram. (6)

0ptlml

Fiber

/\

Post lnline Pre

-Amptifiu Am$ifier Am$ifiu

Tunable '
h- $mtrt
-{
Surce
 .:-E*€.i,"ffi!.,
s
.iF

trqnsmitting en_C and one photodetectot at the rebeiving end, ,' iI

I
To find the optical bandwidth corresponding to a particular spectral t,
width in these regrons, rye use the relationship Q=Iy, which relaics the l

wavelength fu to the carrier frequency y, u&ere C is the speed of light. '

t
r^rt =(#) ta )t l.
i;

Where ttre derivation in fiequency A y conesponds the wavelengilr

deviation A l, around as.

Since the specral wiitth of ahigh-qualiry source occupies only a

narrow optical bandwidth, the two low-loss windows provide many.
additional operating regions. The integrities oftlre independent messages
fro meach source are maintained for subsequent conversion to electric
' signals at the receiving end.
Since WDM is essenfially frequency-division multipiexing at optical
carrier frequencies. The literature often uses the term dense WDM, in
contrast to conventhnal or regular rtrDM. A ksy ftature of rfrDM is that the
discrete warrciengts fironr an orthogonal setof cariers tlrqt can be separated
ro1ry$ and swie*red wi$CI$ intererinE wifr each ofrer. The performance of
acfvedevicescanbeconrolledelectonically,ttrerebyprovidingalargedegree
ofnetwor flexibility. Actirre wDM componens include tunable opical filrers,

A multiplexer is needed to 6ombine these optical outputs into a

Serial spectrum of closely.spaced wavelength sfgnals and couple them
onto a single fiber. At the receiving end,demultiplexer is requiredto separate
a
- the optical signals into appropliate detection channels for signal processing.
' ,, Sinco the optical signail that are combined generally do not einit
any significant amountofoptical power outside ofthe designated channels
F spectral width, interchannel crosstalk factors ani relatively unimportant
at the @nsmitting end.
15. (b) (ii). Give a brief acdount of the different types of losses to
r be considered in the design of an optical link (6)
F
i
Ans: An optical power loss model is analysed. The capital power
received at the photodelector depends on the amount of light coupled
 Oe,ticqtCommunicattonGhiag0 .,, -. ' ' ''?7' .. -
'&

ffiiftr and the losses ocpuring in the fiber at the conneclors and
bplices. Baoh bf these loss elemerits,is eipressed in dB as

The lirnk lossbudgetsimptv'ionsiders{eltttr*f.?tl9:1f*'en

, trre source and ttephoiod*ector, and allocateg
tigti this loss
".:lbt:
"
;#;;r, connector loss, splice and system margin' Jhus if Pt is
loss attached to
tfr. opti."f po*., emerging iom the end of a fib'er f,lVlead , :

the ligtrt source,

= 2t + o fL + syStem margln. :
.'..Thesplieelossisineorporatedintothecablelossforsimplicity.
SONET
15. (b) (iiii Oraw and explain the basic fotmat ofaqSTS'N

the next step in

Ans: with the advpnt of,fibreoptic nansmission lines,
lhe evolution of the digital time-division-multiplexing sche19 was a
foniat calledsynchronous'odtical network (SpNETI
'' rtrna*a.rignal .

CpmmonlyusedsoNETandSDHtransmissi9ryates: - r r rl r t I

Electrical Level Line Rate (Mb/s) SDH Equivalent

oc-1 STs- l '51.84

oc:3 STS .3
i

I55.52
I

srM-r
;;;
i
!

STS - 12
I
STM.4
oc-12
STS -24 1244.16 STM.8
oc -24
'oc-48 STS - 48 2488.32 STM. 16
', STM -32
oc-96 STS.96 4976.64

oc -192 , strs - 192 srM -,64,

90 x N Columns

0 Rows

i 3xN
Cotumns " ^
ETxN0olumns
Fig. : Baiicloimat of an $'rS - N SQNET frame.

C
 i - -E-%;-

t'

K 4265
B.E'B.TECH. DEGREE $(AMINATION, MAY'JUNE 2009.
Seventh Semester
Electronics and Cdmmunlcation Engineering
EG 1402-.OPTIGAL GOTIJI MUNICATION
(Gommon to B.E. (Part.Time) Sixth Semester - Regulation 2005)
,1
.i

*- (Regulation 2004)
-
Tirne : Three Houru Maximum : 100 Marks
') AnswerAllQuectlons
PARTA{l0x 2=201
l. Delirle acceptance angle and crltical ingte of'the flber.
Ans: Acceptance Angle: The:murimum angle 'Q,*,' with which a
ray of light can enter through the entance end of the fiber and still be,
totally internally reflected.is called acceptance angle of the fiber.
Crltical Angle: If the angle of incidence $, is increase a point wil
eventually be reached where t
liilht rti.y in air is parallel to the glas
' surface. firis is known as critical single.
:

2;,Whatisflberblrefrlngenceandtlberbeatlength?-
' 'Aor,, tr'iner Blrefringence: Imperfectlons in the fiber are commor
such as asymmetrical lateral stess, non-circular imperfeit variations ol
refractive index profile..These imperfections break the circulat
symrhetery of ideal fiber and mode propagate with diffq€nt ptase velocity
' and the different between their Refractive index is called'fiber
bireftingence

Fiber Beat Length: If light,is injected intp the fiber so that both modes
are excited, then one will, be delay'ed in phase relative to the oiher as
they propagate. When this phase difference is an integral rnultiple of
Zn,the two modes vyill beat at this point and the input polarisation state
will be reproduced.
. 3. A'100 km fiber is used in a communication system. The
fiber has.3.0 dBltcm loss.'What will he the output power,
when the input power fed at the input fiber is 500 [iW?
a.

t.,a
.
 -.*?tirFj4-' rEEr.
.,q
qF

pi,(dBm)= totog t-1,(w)l

lffi)
-
Isooxlo{ w1
l-l
ru rog
L txl o4w J

=;30 dBM
'.,
l n,onl
P**=r0ros
lffi) !

ln wl
= t0tog
lffi)-.d,z

4. \ilhat is the need for mode coupling in optical fiber?

Ans: The rnode couphlrg will increase the distortion less rapidly after a
certain inital length of fibre due to this mode coupling and diffbrentiel

)n a LED has 2 V applied to its terminalsr-it diaws 100

LED *oo".rsion efficiency from eletrical to optical power?

Ans: Given: Input to'LED = )Y, I = t00,mA, Ouhut Power = 2 mW
o;

LED Conversion efficiency

_+

Ptn=! x 100 x 10'3

f=*
 2x l0-3
)r*
0.2

.6'ltytatistheprincipleotoicraiionoflASrR?
Ans: The principle of oporation of LASER is population inversion the
most photons incident on the system.
7. ""Deflne quantufn effiiency 'and re$ponsitivity Qf'/
:,

I
I
Ans: r The internal quantum elliciency is th-y fraction oflfe el.ecton-
t
l_
hole pairs that recombin! diativefl. If the radiative relombinalion ratl
r is R'and the.non-radiative recomtination ratios ft,,, then the inlernal
I
E

F quantum efficiency is,the ratiolofthe radiative recomUination rate to the

r

recombination ratei
r,

B.

F
8., , A siticon photodiode hrs a'qutntum efficiency sf 65% at a
wavelength. of 900 nm. If 0.5 pN optical power produces a
F
multiptied photocurrent of 10 F,A. Determine its primary
r
I Gi : Q =650/o, L=900 tttttrf; =10p/, p=0'.5 tdf, Io=? M =?
+

Io
o.6s!-
l.mr^ .

9.Defrnemodalnoiseandmodep*rtitionnoise.
Ans:. Modal Noise: It arises yhen the light from a coherent laser is
coupied into a multimode fiber operating at 400 MbPs pnd higher. It
mainly occurs dug to mechanical vibrations and fluctuations in the
frequency of optical source.
+.
l

l6 :a.
 t,i*
E
B.E./B.Tech Degree Examination (May/J'une 2009)

The Partftion Noise: The rnodc partition noise is associated with

intensity fluctuation's in thc longitudinal nrodes ofa lasgr diode. It becomes
more pronounced for the higher bit ra'tes.
10. What are the advantages of using soliton signals throu$h

Ans: Group velocily dispersion causes most pulses to broader in time

as they propogate through an optical fiber. However a particular pulse
shape-known as a soliton makes advantages of non-linear effect's in
silica; nartiSular self-phase mgdulation.
'
: ,

B*t5 x i6 = B0l
PART
1l.(a)(i) Derive an expression to determine the modes
propagating ins lep index liber. (11)
Ans:
,J
A standard mathematical procedure for solving equation is to use
the sdparation-ofvariable's metho{ which assume's the solution of the
fonn
E2 * A: rr(0)
"(0
As ffas already assu$ed

dni' r'
4'(0) = '

The constant y crn be positive or negative, but it must be an integqr.

Since the fields must be periodic in 0 with a period of 2n
'fu

ry*19l
0r' r0r
*ln,-*18
,') I =o
L'.
whic,h is a well known different equation for Bessel functions. For the
configuration of a st'ep-indEx fiber we consider homogenous core. of
refractive index n, and radius a, which is slurounded by an infinite

The modes in step-index fibre all modes are hybrid mode except
hichy.=0,Thuswheny:0....'.
' ,J ,
r(ua) , K, (ot. ) ^
-w !
 -:r@r*lQq!,
,. "*
,.s,

ical Comrnunication (EC-1 402

|

:
,

which corresponds to TE,,,, lnodes (E., 0)

-!

- t_

L ,,
"
ulo @a) oKn (ro, )
whiqh coffesponds,to TM
omodes

' '
F.

.' Fig.
Variations of the Bessel function S,I*) for I (e) orders
!V=0, I ,Z)

(u + w) a2=(T)'bi -.i)
-

i :'./-

t-

r p/1c TMOl

EHO.

-EH21 \
!HE22
2'1 2 3 4 5 6 !
 -,.sFg""',*il%G.
.F
e.

B.E./B.Tech Deqree Examination (May/Ju1q@qg) ',,,

r
t l.(a)(ii) Calculite thetumerical apeqture, cut-olf Paramefer and
I
. sqpported by a fiber-having Fr @:dre\
number*of modes sql
= 1.S4, Fz (cladding) : 1.5, erore radiut 25 pm and

':

N, = (n,' -- nr')'o
.,4=(1.54F-(1.5))'o:
i)
-
ii) Cut:off Parameter

lft =' WNAz

rr r2
(^,

2 x 3.14 x ?5 xt10-6 x 0.348

l/= 2 n;al\rlf, =
1300'x 10-e
\

iii) Number ofModes:

p = (42.0276)*

ffi tl = 65%
!l'

l, = 900 nm
Iu= 10 vA
P =' 0.5 vN
I,: ?
M.?
 -.:-4idr''*twf,x.
.-iiE
S.

Otpticql Communication (EC.1 402)

[r[ =lJL
.IP n

Q='-aa:
tqP,,

F
I, 6,626v1S:'rr2 x3x,l08
0.65= ffi-r, 'i
x
900x l0-e
.

0.65x:l .6x l0-re x900x 10-e

rr,=
! 6.626110-t? x3xl08

'Ii=4.78 x 10-?3
,.

E* r
,
M_.M

fifi--
t l0x10-6
=:, 4.?8xlo-23
-

lv{=2.092
-,
x l0r7

lr
The cut-off conditions for some lower-order modes
t Cut-off :

Jo@a) = Q

Jr(ua) =0

a'o)' k)' -ni(3 I

-\
i
r2 o?-ol
i
'.: ?qGilEFFEFE=:.4.ffi3%..
:
'&
8 . B.E./B.Tech Degree Examination (May/June 2009) s .

-, ' .' . .ryf E n?, the HE,, modc

has,no cut-offand to only
is zero
"eaibt "Xist
when the'core diameter

A ray congruence will be accepted by the fiber

: ,i '
: ^.ilz
:'::l

. 1L O) What are the various'fettur.gq of grrded'index flber? '

Explein the refractive index profile and ray

tra[spisslon in a multimode-graded in{ex liber. (16)
Ansr Multimode Index,Fibre: :

. lUe propagation of light along a waveguide can be desuibed in

z.
J.
- terms of a:set of a set of gqided electromagnetic waves called
'

'' o These guided mode's are referred to as the bound or tapped

i: , :

o Althougb many &fferent configuration's of the optical waveguide

has been discpssed in the iiterohere, the most widely accepted .

:sfiuctedisthe singlesoliddieiectiioylinderofradiusaandindex.,

I
 -.qp".,,.*,W*r.
, '.;$
B,

QplLc-ellqmmu n ication (E C- 1 402)

Buffer Coating

Fig. : A Clrcular $olid Core

. ,Irr a n vatvflt!.t. vrtttrE
.... f
.

Variations in the material-composition of the core givg rise to the 2

commonly used fiber. The RefraCtive Index of the lore is uniform
throughout and undergoes anit abrupt change. This is called a -step-
index fiber.
The core refractive index is madeto vary as a function of the radial
distance from the 0enter of the fiber. This:type is a graded inex fiber.
.

Ttrey are further divided into multimedia index fiber.

'i
-

/m,
aaaalaaaaaaat
t=a $ t25 - l4o qm
t

t=o
a.aoaaaaaaaa.t

t so-1ooqm
Multimode
lndex Fibre

: Fig: /

125-140 vm Ctadding

50-100 v,ir -+; Core. .

A disadvantage of multimode fibers is that they- suffer from

intermodal dispersion. This msans that the rnodes in a given opiical pulse
arrive at fiber end,
Scattering Lossesr Scattering losses in glass arise from'micrgscopic
variation's in the material density from compqsitional fluctuations and
from structriral unhomogeneities or defects orcuiing during fiber
manufacturerln addition source glass is made up of several oxides such
 .re:*ffir,
.,n
s.
.
$:
B,E./B.Tech D ree Examination (MaylJune 20Qg

--scut - 3dt t' ,

0r.., = p'K BT.f

Br
trn=
, where 'p' is the photo elastic co-efficient

8zc3
cr=
#(1.')uav
(an Y
l-l
(6n')'
' \ap)

,b F

O .+.
'' lr-
4., L.
qo
SE I
trm I

9e I
I I
I
I

l I
I
a
I l1
I I I a o
a
l1 I
I I I
l1
I
I I I ,. a
!
I II ! I I
a

Wavetength (nm)
Fig.
--
0u, =ggtlEo

Bending lbsses: Microscopic bendS having radii that are large

rcr diameter, Rando
- Microscopil the
lendt :.f
can gdse urhen the fibers are incorporated iryo cableq.
{ber axis that
nes: Ifthe influence ofmodel coupling is ignored,
the loss for a mode of order (v, m) for a step index waveguide is

c[
-vm'PP

ctvm.!*

o(0 =
 -.rqp",,.'%l[8.
\.. . r;*
R.
'i

cal Csmrnunicafion (EC-l 402

I2. (a) , ,' What are the losses, on signal attenuation mechanisms

Ans: Attenustion: Attenuations of a light propagates along a fiber it

ls an rmportant..Consi{eration in ,the design of an optical computer
system. .'t

Attenuation units: As light travels along a fiber, its powerdecreas€s

exponentially with distance ofP(O) is the optica! power in a fiber at the

P(z);: P(0) e-o,'

t. t-P(o)l
or=z"Lrii]
is the fiber co-efficient given in units ofkmt ' I '

{ lo. lrroll
_tosl"(r)J
a(dBtkm) =

, " = 4.343 o, (km.r).

Absorption: Absorption is caused by three different mechanisms.
. 1. Absolption by atomiq defects in gtass composition.
2. Extrinsic absorption by imptrity atom's.
g

3.Intrinsicabsorptionbythebasicconstituentptorn's.

o iot to' to: td :

Flg.,: Tlme after lrradiatloruts)

The dominant absorptio{, factor in silica fibers is the piesence of

minute'quantities of impurities in the fibre rnaterial. These impurities
include (OE") ions. Transmission metal impurity level were ar-ound I
part per Million

tr
 -: rEE,*r@f,x.
'. *r
R.

12 B.E,/B.Tech Degr:ee Examination {Ma$une 20091

*

tr
O
aJ
r* L.
E.E
tr6
Sg

Wavelength (nm)
Peak Fiber 'I

Fig.
I
I

ttre *ilheA curve is theattenuation foi{ow water-peak fiber,

Ls4.z7 - /+.og\
= ;ee;ftoxto-'exn[fJ
^-,
, ,/
'
should be located at the focal point ofthe lens, '

. n n' n'-n ,t

---=-
. I , -. . t

+ Lighttravelsfrom'lefttoright.
.-+ .Objectdistances are meadurgd as *ve tothe verture and negative
to the right.
+ , Image distances are measured a positive to,thrc Right of a vertex
and r-ve to the left.

LED

,s
aar

F!9.
 ---/. s,
tical Communication (EC-1 4A2

I
F
;,

t
F

lr
;
i

qt

t"

i
I

i
 -.-?di*,1-"r
"t]WAg,
.'.F
s.
s
B.E./B.Tech D Examinatlon Ma /June 2009

I clr
Total DispersionP-' t{

=l x 50
: 50 n sec/km.

13. (a) (i) Explain thq rt:p: involved in spticing the flber. Discuss
the various splicing techniqUes employed'belween two -

fibers. (8)
Ans: A fiber splice is a permanent or a semi-permanentjoint outbetween
'fiilro fibers, orie must take into acocunt the geometrical different in the
two fibers, fiber misalignments at the jout and the mechanical shength

Spllcing technique: Fiber splicing technique include the fusion splice,

the g;groove mechanical spice and the elastic tube splice
so'se.
The first
techniqr. yields a pennanentjoint. Fusion Splices are made by thennally
bonding.together proposed fiber ends.
, ' Fusion Spticlng,of'Optical Flber

Elec{rical

.a

Fig.

13.(a)(tt) Explain the lensing schemes used to irpprove optical

Ans: Lensing Schemes to improve Optical Source fibrel If the

source-emitting area is larger than the fiber. Core area then the resulting
optical fibel coupled into the fibers is the maximum that can be achieved.
:

Th9 function of the microlens is to magnify the emitting area to the

source to the match the core area of fiber and face exactly.
:
 .,*
S.
-ss
Optical Contmunication (EC-1 402)

Non-lrnaginary Nlicrospherc: One of ttrc most cfficient lenling

methods is ttre use o'f a non-imaging microsphgtre, It use"s for a surface
erpittcr is known.
To collimate the output from the LED, the emittihg purf4ce.
V-Groove Optical -

ii.

a^.

r
c*r

Epoticd or
Clampod Hen

' lll
Ftg.

Alignment Concept for an Ellstic,Tgbel

Capillary
.',Tube Slze

..a.-
Fig.
-Splicing Singte-mode: The loss depends on the shape ofthe proPagating
62
mode, fur Gayssian shaggd of the loss betwe.en identical fibers is

,-ll

for"angular misalignments in single mode'fiber$, the loss at awavelength

l, is 62

E -10 log

ffi

m
 *:- [x
, '.,x
E}

16
^.

B.E./B.Tech Degree Examination (May/June 2009)

*,,
(or)
\,
(i) Explain the basic LED conligurations used as optical
13. (b)
,our.u. Derive the expression
-- for quantum bmciency
-d and opthal power generated in LED's. (10)
Ans: Basic LED Configuration: I ED configurations such as
homojunctions and single,&rd double heterojunctions have been widely
investigated. The most effective of these stnlctures is the configuration,
This is referred to as a double-hbterostnrcture because of the two
different alloy layer's one ach side ofthe active region. This configuration
evolved from studips on laser diode.s. By means of this sandwich
structure of differently.composed alloy layers, both the carriers and the
optical field'are confined in the central activE lqyer. The baird gap
differences of adjacent layers confine the optical field to the central
active layer.
Expression for quantum .tfili*ncy and optlcal power: fui'9*rrs
of electrons and holes in P and n-t5pe material. The excess densities of
electron n and holes i' are equal, since,
ra
o
B
&
*,
EI
cr
ar, \
3 r..Qurfa1;5*itting
o
o \
ara

-g
(1r'
u
Fig. :

Two main causes in Intramodal: Material dispersion c.auses due to

the variations of the Refractve index of'the csre material as function -_a

of wavelehgth. Material dispersion also re referred to a chrornatie

dispersion. This causes a wavelengthdependence of the group velocity
of a given mode. r r
Waveguide dispersion: This causes pulse spreading because only part
of the optical power propagation along a fiber is confined to the core.
Polarization mode Dispersion: The light-signal energy at a given
wavelength iii a single mode fiber actually occupies two orthogonal
polarisation states.
 _..={isr"_.wf,s
..9
s.
Optical Communicatiori (EC-1 4A2 17
re-

13.(b)(ii) Explain*the modulation process inovlved in LED and

discuss its frequency response. (6)
Ans: The.transmission of digital signals, there are many potential
applications for analog links-operating in the multigigahertz region. This
is defined aS the ratio of the mean-square signal current to the mean-
,.
square nolse current.
.l-tJ

;
!

IUS
p(r)=f,tt a-
+ms(r)l

Tri --
I8
here AI is the variatioR'ls current abbunt the bias point. In order not to
introduce distonion into the optical signal. The modulution must be confined
to the llnearregtonofthe light ro** output c-urue for expgnential decay of
excess carriers; the radiative recombination lifetime is

where the bulk recombination lifetime t is

'T Tr
1r,

In general r, and \ n, are comparable for direct band gap

. semiconductorp such as GaAIAS and InGaA^Sp. Thus quantum
a

efficiencies of 60-8004.
r current injected i$o the LED is /,, then the total number of
recombinations pr seiond is-
R'*ftn"= I/q ,

. R, = rl ,ni I/q.

P,n,
-
- tl ii, I, ,hr:
I.
Ii,i, :hr=Iint ---
hc
(o)=m 4nrn,

rP=t.r
- ,l ,*, rPtoo := =tln'(n+
'f*t 1)2 =

I
 sl.

* i

l-
o
B
t
o
o.
*r,
3
o.
*.4
3 I
{ ,'r'
t.
' Bias Point
o I I

-.o
ru
I
l
a
ta
I
t" t+ Modulation Pf
o I
Input Drive Current

Fig.
of 191-
The excess carriers can recombine either radiatively
is
,aaui*tr. tn *aigive recombination a photon of ensrgy hy,which

dnJn

The' equilibrium condition

,e

Jt
ga

The internal quantum efficiency in the active

Region is the fraction

R,
--!
 ':ia.

.F
B,

tioat Communication (EC-1404

19

')
'l
photodiode and Avalanche photodiode. (16) ,''' ,

Ans: PlNPhotodiode:When an inciclent photon has an energy greater

than or equal to the band-gap energy of the semi-co"Ar.ror"*lierial,
the photon can giv.e up its energy *JLxcite an elecffon from the uul.nce
band to the conductiont band.
hqn"l

Load
eeistor
Output

Fig.
U

This process generates mobile elechon hole pairs. These electrons

and holes are known as photocarriers. :

EnergYJBand Diagranr u

i
I
Gap I
I.
I
.l
I

I Valence Band
Depletion Region
:
.t
Flg.
-t

t- -
Explain the noise S;1.., and distur,bances in the
14.(b) (i)
detection of optical pulse in the rbceiv;.. tot ,--
Ans: The term noise is used customarily to describe unwanted
components of an elechjc signals that tend to disturb the toxin and
the signal in a physical system. We shall concern*J;.irly
ry-essing,of
with internal to the System, represents a basic limitation on the Txion or
deection of signals.

,. tZ
tr-
€
 .:rry: "r-
[U.
.F
s.

B.E./B.Tech ree Examination MavlJune 2009

-t. qE
ff=+ I pQ)dt =€-
hy'
,TT d

e power according to poisson distribution

F'e*F
n!
give
The random nature of the avalanche multiplication process

(r-J-l (1 *rL1
t M)
before it eniers the input circuitry of the following amtri- In'order
for
carrier multiplication to take place, the photogenerated carriers must
travers a?egion whers'a very high elletric field.is pret.nt
The cornmonly used stnrcture for achieving carrier rnultipliyflon *{h
.

ach-through constnrction.r The teach-

APD) is composed of a high resistivity
p:type mateiial deposited as an epitaxial layer on a p* strbslrate

.s
G
o
<)*

tr
o
ts
J
()
to$
,&-

1m 200 300 4m

APP is characteriyd by its responsivrty

: Thus the performancsofan
RAPD which isgiv*n bI, Er*, = ;i l1t! =tsM
Photo AbsorPtfon Co-eflicient.
, P(x) : P- exp (*c*f)
If _ g is expressed in units of Jlectroo volts (ev) then 1..
__ E",
is given i4
units of micrometers
i' hc
,av l

r ' ,. 1-----\ ;
A.([rm) = E
I
 ..re:"%*f
.F

.
s.
tical Communication EC- fiAl
I

1.24

,l

If the depletion region has a width W, the total power absorbed in

the distarrce N is
(D

p P(tt): - IorP, exP(-orl)d*Pn

- t
(1
w
- r-o''o
)
il j." 0
;"*

.l -- photodiotle
R" is the entrance face of r--
.t

., ,P
I

'

I
number of electron holes pairs generated
Tl=
number of incident photons

' Inlg '' "

Tl=
P,, lh^t - '

Avalanche Photodiodes: Avalanche Photodiodes internally multiply

,.
, For exponential decay of excess carrier the radiative recombinarion

t_"

lifetime is l'lin, = I+Tr lTn,= -\r

--
wherethebulkrGcornbinationlifetimeTis'
; -.

111 I

,i?
T Tr 7n,

, In gBneral \ r and r n, are comparable for direct band gap

,/

semiconductors such as GaA I As and InGaAsp. Thus quantum

If the current'inJ..i.A into the LED is I, then he total irumber of

rccopbinations per second is
'l
R, +&, =i
g
 .--*,gra', "a@f,lF,
's.
.-lr

May/June ?009

R, = tlio,f I q ,Noting that R,. is the total number of photons

. , 14 4nrn,
Pin, :Qin,J / q h:r.= f(0) =
1'1,"
;t, Gffi
zP=tr
'I an-P.
tnt= f*'
n( n + t)2
{,
I4.(bxti) Derlve tho expresslon for slgnal-to-noise ratio .

'
.1 obtatned at the oytput of rn optical recoiver' (10)
t The range from individual.If(t)z voiee chaniels tp microwave
Ar.l
lucks operatirt in tne multigigahertz Region. Performance fildelity
is
*."t"tua in teilrs ofsignatiinoise ratio. fnit it atfined as thc ratio of
thg mean square""-.ffi:I,tixil,.'r,," noise current, :

P,.is the average transniitted optical powel: B(t) is-the.analog

moduiation sl*r,*O 'z' is the modulatior.r signal index defined by
AI
j . ffl= t
" lt
N -+ is the vari*ionin cirrent about the bias point. At the receiver
end, the photocurrent generated b the analog optical signal is

,j
: /r(t) = &l'lPJl + zcs(t)l,
Erplaln ttre ampliffcaaon Pghroiiro ln EDIA" Discuss the
F-
&
p*
Y 15. (a) 0
possible copltguradons of EDFA rvttlr.neat 4qgrrlu (lq
F
8:
f;l

Ans: EDFA: The operating regions of these devices depend on the

host material and thp dopi4g elements. A popular material for long-hand
telecommunication applications is a silica fiber doped wrfh erbuim.
"
Amplillcatlon Mechanlsm: External current injectors to excite'
.r."'*ip',o hig[er energy levels, opticai fibpr amplifiels.use :ptical
purnpini. In this process one-us.es photon to directly'raise olectron is
elevated must lie generated.
The erbuim atoms in'Silica are E3* ions, which are erbuirn atoms
,U* iurOiort tlree of ttreir outer elecrons. The trro principal levels for .

telecommutication applications are a metastable level

 -.rcffr,*%f,E,
.F
B.
*'
23

lnput Drive Current

iNz = 2q(1, + Ir) VzF(MBe + ZqI * Be * ff t te{

b
rr
Ip -+ Primary photocurrent
F
T
ID -+ .Primary bulk dark current
'{_ :

,
7
r -+ Surface-leakage current
F'ftr) -+
Excise photodiode noise factors
Be +
Effective Receiver noise B,W
Req +
Equation Resistance of photodetector load and amp.
f, -> Noise figure of the baseband amplifier.
i: ttz(RM,,n),
- ii
Art'
-;|-
d/rY
\q(Rp, * I o)M'F (f W)) .ae + (4KrTB,/ R.o )4
tl2(I ,M *\'
Zq(I ,+ Io)M'r(M)4 +(4&TB,tR,u)4

mz I, m2.RPr
S/N=
ffi 4qBe
15.(a)(iD Dqrive the expresslon'for EDFA question conversion

[:
F
Ans: The amplifigr gaineventually starts to saturage. The reduction of
it

gain in an EDFA occurs when the population inversion is reduced

k'
i
t:
i.

i^,
.
l--r-,r,
srtr+ L,
Pro* s P",, + pr.,n

- ,rr.r-.r, ;,;
F
i,r-
t where P*. * the input power and Lr, l,; arb the pump and signal
[.
:".
wavelengths. The fundamental physical principle here is theamount of
E signal energy that can be extracted from an EDFA can't exceed the
E
F
*' pump.E', prr*
4,ut*4i, e_sLsr
F t.-=Voin l.r
?n- B,t,lB,Tech Degree Examination (lviaylJune 2009

[.". Qr,, is unitY]

E
6
t,
tr
G

;
o
,n
o

0 1480 1500 1520 1540

o The bottom ofthe f r,rometastable band separated from the bottom

y 0.814 eV. This energy

, top of the olrrr.ground state band by about 0.775 ex.

T,he means that possible 980-1480 run. The decay phenomenon is known
as spontaneous emission and adds to the amplifier noise.
 ..]]s#.....:,....l,:.... -.-=l{.itrr*a@[W
,..F
B
a
L- ,r tr.! ' i.
t.

tS.(U)(i) Explain t'he basic franre structure and netrvork

' architecture of SONET. (8) aa A -h. r v-tm /f\\

Ans: The SONET forrnat callecl synchronous optical network. The

.physical layer aspects of SONET/SDH as they related to optical

Thnsmission Formats and Speed

: After man'y differences of opinion of knplementation phulosopy
. were resolved resulted in series of ANSI fi. 105 Stturdardsz for SONET
and F series of ITU-T recommendations for SD[P3.
: Astandard SoNBTt6rminology a sectionconsists adjacent pieces
. ofequipment; line is a lo:rger link that connect's two SONET devices.
a
and apath is a complete end-to-end connection, of the eight possible .

combinations of rin! types, the following trvo architecturphave become

popular for SONE-T and.So, networks.

c
tIEI
t((
ro
Eol Sectior €o syn-
& l-ine
}J
3\ Over
t6 chronous
tr Payload
ot Haad '.tr
Iaj
IG Envelope
s\ I
\-a/'{ V
3 Columns 1 Column

'i

.twofibprounidirectionaI,pathSwitched'ring
.
.\
o two-fiber or four fibre, bi-directional,'line.
,,1
.,
a---
The common abbreviations. of these configurationg are given in
 '*r- r'
,.F
s.
B.E./B.Tech Degree Exdmination (May/June z00g
Architecture of two-liber unidirectional:

Primary Path

Llnear, Chain

Network SONET

_'

Thus the basic frame structure. ,l

ls.(b)(iii Explain the operetion of wDM components. (s)

Ans: A characteristic of WDM is that the discrete wavelengths form
an orthogonal setofcarriers which canir rrp;;il,;";ild *itched
wittrout interfering with each other. This isolation between channels holds
is kept sufliciently
, low to
prevent non-linear effects.

mM Newtorks requires a
variety ofpassive and active dwices to combine,d istribute isolate aud
dqurv optical power at different wavelengrhs.

-/
 .,x
B.

tical Communication

L,

l.?

ln-Line Arnp

Reoovgrs ,

Tunable $ources

Frgquency deviation A, corresponds to the wavelength Al, and l,

.

A
t-
I-
.)d
-
m
13
-o
\r,
l-
't-

- (u
=
e
ff
0)
fl

'-
14THZ 15 THZ
Gvelength (nm)
Fig.
The implementation of passive and active cornponents in a typical
WDIvI link containing various types.of optical amplifrers, At the
nansrrnitting end thepe ** r.*eral independentiy modulated light Sources,
each,ernitting signals at a unique wavglength, here a multiplexir is needed
to combine these optical oututs into a continuous spectrum of signals
and Couple'them onto a single fiber. At the reeeiving end a demultiplexes
is required to separate the optical signals into, appropriate detection
channels. !.
 ,

JF
B,
s
i*

B.E./B.TECh. DEGREE EXAM t NATION,

NOVEMBER/DECEMBER 2009.

(Regulation2oo4)

EC 1 402 - OPTICAL COMMUNICATION

(Common to B. E. (Part-Tlme) Slxth Semester Beg u latlon 2005)

TlqelTirc,hours.', ' ,'. , " ,Madmrm:l00"rierls

AnerverALLquestlonl.
I PABTA-(10x2=20markr)
' 1. Whatlstunnelefrecttn Opdcalftbci?
Ans:'. ,]
' ,.
':
The leaky modgs are continuously radiating their pbwer 9ul of
, the core as they propagate qlorrg the fibro. This:power radiation
out of the uravegBide in qpantum mechanipal p[renomenon is
referred to as'tturnql effect'. , .

Ans:
Skslv ra),$ n.ritranspitted thfqqS}. thc -fibre axis. The
"re
skew rays follow a helica! Path in the optical fibre.',It is very
dfficult totrack'thsskewray$as thcy donot lic in asingleplane.
3. Vtrhat arc mlcro bends? Itrow they lre formed? '
,t
:

Micro,bends are repetitivol small: scale fluctuations in. thq

radius of curvature of.the fibre axisl They are formed either by
non-uniformifies in the manufacturq :of'' ihe' fibte or by non'
I uniform lateral pre$ures created during cabling of fibers.
4. A Muttlmedta GI fiber cxhlblts a totaf pulse broadenlng of
0rl psec over a distance of 15 km'. Es{mrte mardmrq posslbte
Bandwidth on the line.
 ."EF:*til--.

.F
s.

, Ans: ' , .-' '.', / .,.

. ' Bwr = Br = $/4) = (l/2 x 0.1 x 10-er; SlMHr.
5. Ilcftne tte threenodesof the cavlty ofLaser. .

.,
Lomgltudinal rnodels are related to the,length Lof the cavity
- Latieral modes' tie in tho plane of ,n" plrl',junction. Those
modes depend upon the side Wall preparation and width of the

Transverse modes are associatcd \dth the Blectro Magnetic

field and beam profile in the direction,perpcndiculal to the planp
of the PN iunction. These.tmodes dptermine the radiation pattern
of the lasgr.
6. What is a DFB Laser? What ls lts maln dllference from the

Ans:
In DFB Lascr, .the lasing aition is obtained by periodic
variations of refractive index, which are incorporated into
multilayer structtro aiong'the lengh of'the diode. DFB Laser
does. not require gpti0al feedbaek trnlike the other lasors;:
1. r,m oui tm''vrhres or.Ot*rd6ig ltavelengthg and Respon-
sividce of Sl, Gc and InGa As photodiodcs.

R'esponsivity =,0.4 -t.6

. Ge:800 - 1650nm & 0.4 - 0:5
InGaAs: lfm - 1700nm & 0.75 - 0.95,
t. \ilhat is meent by G/0 notse corncr flequency?

. The (1/f) noise corner frcquorcy fc is defincd as the frequency

at which (1/f) noise, which domi{rates the FET noise at low

[,
 ."=EcFr=-4h.
..8
g.

Oplleal Gommunlcatlon (ES- t *0e)

*
t'

frequencies and has (1/f) power spectrum becomes equal to the

A Laser which oscillates in a single longitudinal mode under

CW operation may experience dynamic line broadening wfen the
injection current is directly modulated. This line broadening is a
friqueney chirp associated with modulation induced changes in
the carrier density.
i0, lryhat are the technlque used for mtnlmtziqg reflecfion nolse?
Ans: ,

rclative to the laser emitting facet. fhis diregts {eflected light

away from the fibre axis, so it does not re-enter thc waveguide.
il) Use index-matching oil or gel at air-glass interface.
tii) Use connectors in urhich end faces make physical contact.
lv) Use optical isolators within the laser transmitter rnodule.

PARTB"(S x lQ = 80 marks)
I . rt /i tt
11. (a) (i) Wtth the help of sultable diagrams, explatn the followlng
eoneePts tn Opticql fiber transmlssisn. (12)

(3)mode coupllng.
Ans: Evanescent Fleld

Cladding
Guide

\r1hve vactor or the

incident plane wave
$tanding Vhve

Flg, 1: Evaneacent Fleld

f,
+ ..*
g,

4 B. E /B.Ttch. Drgrer Exrmlnrtlron (NovL/Drt. t003J

Mrtuhl Reflecting Plene

Reflecting
lnterface

Lateral Shift

Flg.'Z: Go0+Haenchen llald

F is : H a v' n b I v I i be r pe r-
"#Jl$;:',il'ffi i.tr $Ilfl :[: :', lfi ,ffi:' :

iregularitytthe oor€-claddirrg interface; (b) llber bend

(a)
The incident roy, the norrnal to the interface, and the
reflected ray all lie in the same plane, which is perpendicular to
the interface'plane between the two materials. This is called as
plane of incidenc€. When light traveling in a certain medium is
reflected off, an optically denser material. This process is referred
as cxternel reflection. Conversely, thc reflectign of light off:of less
optically dense material is chtleOls internai reRection]

F
Effi'fET
..*
g.

Optlcal Communleatlon (EC- I 402) 5'

11. (a) (ii) A Multimode GI frber has an acceptanee angle in air of

refractive index at the core.axis is 1.52. (4)

IllrEli,l ED Dr'

Ans: ,

, Acceptance angle '+

-t

Lir' - sin-'(Nn) i'e" NA = sin 0s

NA = sin 80 = 0.139

NA = nL (2 L)ta i.e., A =
1
; (ff) '
a
1 /0.139\2 ^ aa,r
A=* !;=rl
; [,*] = Q.0042" or A'MVo' ]

11. (b) (i) Bxplain with sirnple ray diagrams 0) the multimode Sl
fiber and (2) the single rnode SI fiber; Compare the
advantages and dfSdvantages of them trro S1rys. (10)
Ans:
,
Step Index -Single Mode Stcp Index-Multimode

Here only one rnode isHere more than two modes

, 2,
are possible.,
3. Since the onC mode is present 3. It can carry more number
only one information is passed.
4, If the data is loss then there is 4, If the data is loss theq it
no replace since it contains can be replaced since it
one mode. contains multirnode.
No reflection takes place. 5. Total interval reflection
 11. O) Gi) A slngle mde step index fiber has core and cladding

, Pel1mit tts operltlon over the wavelength rang€ 1.48 and

;
I
i
I
single mode transmission at the wavelength of 1.3 /rm.
r
:

'i-l

', Cord diameter required to allowi between l.48,and 1.60

Core radius'at'

n* n1(za)1" (a
j

=rq?;;? ; '.v = z.4as)

.'=2 znr :

.- -z 7.4952' . ,nn2 0.'0081

(A = 1.498- - Tx tH
1.498 =4-.*#3 = 0.002)

'rr, * 2,4as x I.48 x la*6 lz x 3.'14 x 1.49g ex0.002 )t /2

= 5.98 x 10'6 m
G

o'oz' '
= ),.40s x 1 .60x 10-6 t2x3.14x 1 .4gg(2x0.002 )1/2
' = 6.46 x 10'6 m v

r..=3

' *a3'
= !,.405x1.60x10=6/2x3.14 x l.4gg (2x0. gaz)| ,2
'- - aat
 ,,,-_ffir-l-te$f
,1

*
..E

Opllcel Communlortbn (EQ- I {Oe} .7

12, (axD Descrlbe the llnerr and non-Iinear scattcrlng losscs ln

opdcatllber. (t)
Ans:
scattering loss in glass arise from microscopic variations in tho
material dengity, from compo.sitional fluctuations, and. from
structural intomogenesis (or) aefects occurring during fiber
'manufacture. In,addition, since glass is made up of scveral
oxidcs,
such as SiO2, GeO2 and P205 compositional fluctuations.can
occur. These two effects $ve rise to refractive index variations
which occur within the gtass over distancps. that arc small
compared with the wavelength. These index variations cause a
Rayleigh-t1pe scattering of light. Rayleigh scattering in glass is the
same phenomenon that scattering in the glass is the same
1 phcnomenoru that scsttering light from the sun in the atmosphere,

thercby giving rise to thc blue sky.

' The exprossions
for scattering induced attenuation are fairly
complex owing to the random molecular n4turg and the varioG
oxide constituents of glass. For single-componen'ts glass the
' scattering loss at the t risulting rrom &nsity fructufi;i can bo
approximatedby

Here, z is the refractive index, K6 is the Boltzmann constant,

Bp is the isothgrmal compressibility of the material, ?y is the
temperature at which the density fluctuatcs.

.,'
fabrication ian also cause scattering of the light out of the-fibre.
These defects may be in the fori of ;"#;J;i;;;bil:
. unreacted strating materials, and crystallirca-i"gioffi ti. gt.rr.
I
At .l longer than pm, infrared absorption effects tend to

\
 , r"'jilEiEnr,
.Et.
.-rF

s
B.E./B,Tech. Degrsa E(amlnatlon (Nov./Doc. AOO9)

12.(a) (ii) Silica has an estimated fictive temperature of 1400 K

yith aT isothep$ compressibility of z 10'11 *2N'1. *
The refractive lndex and the photo.elastic coefficient for
Silica are 1,46 and 0.286 respectively. I),etermine the
theoretical attenuation in dB/hm due to ttre fundamental
Rayleigh scattering in Silica at optical wavelengths 850
trDr 13f0 EIII anfl f550 nill. (Boltzmann's coost"ot E
1.3g1 x 10'23 JK-r). (E)
Ans:
The Rayleigh scatting coefficients may be stained for each
wavelength from the following equation.
r
. ,
^-2R
1.895 x 10-28
t

,
F.
$
i.e.rTR=ffim-' -l
r
F
At wavelength of 850 nrns
1.895 x 10-28 -r' !r t i:
Tp=r\ %m 0.363 x 16-3 m-l
GsO
:
10-")'
The transmission due to Rayleigh scattering in is
-{
t r \ I
Attenuation lolosro = 10tog
=
ffi Ed
= 1,.57dB/km.

At Wavelength of f3l0 nm:

yR=-
:" = 0.064x10-3m'r
v'\'\"
(1310'x tgYp
..
,

pxm = bc(- 0.064 x 10-3 x tO3)

Attenuation = l0log
6d
= 0.277 dB/km
 *'Ettf;,i .

.,x
B.

Opllcel Communlcetlon (EC- I 402)

if.

At Wavelength of 1550 nm
ao

yR= ,1'9li
.;':=;i
_ _?3,
^

= 0.032 x to-3 m-l

Pkm = exP(- 0'032 x 10-3 x tO3)

A A,"1

= Q.968

Attenuation = l0rogro
ffi
. = 0.139 dB/km.
(or)
12.(b) odsl putrse broadentng ln Mulfimode
(i) Describe the intermodslp$*
$ and GI fllers and derlve the cxpresslons for _tlelr -

lifirl'ffiffirH',ft
fiber output. (8)
m ffi,#i#$ tl
Ans:

,efract[ve index prolile and raytransmieslon in step lndex

r,fl?rllrt*rrrimode step indextibe6 0) single-rnode step indexliber

I'
 ' "',
].'[email protected]*=, '"a- h,
, .,1{
s.
t0 B.E./B.Tsch. Degrec Exlmlnrtlon (Nov/Drc. zooe)

Refractive lndex (n(r))

i'/ ! \\r
ili u
li \l
F \
i
Radlal
rI distance 0)
Gore
I
Axis
r a

- l lor d ltleront rn lueg ol a

Ii

Fig.: The refractive index profile Fig.: An expanded raydlagram

and raytransmission ln a showing refraction at the
multimode graded index liber various high to low lndex
lnterfaces within a greded
indexfiberi glving an
overall curved ray path

&
 ''-ryfim., **-
E
r -.9

B
Optlcal Gommunlcatlon (EC- I 402) lt *
'
12. O) (tt) The beat length slngle mode opdcal frbcr ls 9 cm
ln a.
when light from an iqiection laser with a spectral line
t
wldth oi oma nd a peak wavetength of 1550 nrn ls

estimate the cohere4ce lingttr ln thls sltuation. Nso,

calculate the differencd,htween the propagation con'
stants for the two orthpgonal modes and check the
resnrlt. (t) ' '.' ' .

Ans:
To find the model birefringence, the following equation

Coherence Length
: 12 -')'
Lur=qm=rJffi0.155. x 10
bc- BfdII L7Z" i0-6 * 1 x 10-e
= 13.968 m
i The differences between the propagation constant for the two
i. orthogonal modes

... (3)

The result can be checked by using equation

2n B^
n
B 'y
'x -S - r
A
T rn-6
2n
= -\]t72l'-7o-" = 69.8 ...(4)
.. 1550 x 10'

13. (a) (i) Erplatn the theory of stimulated emlsslon and laser with
the help of Energr band dtagrams. Also explain: lts
temperahrre dependence characterlsdcs. (12)
 ..F
B.

12 B. E.i B.Tech. D,agrcc Examlnatlon (Nov./D,cc. 2OOg)

Ans:
: r'
Laser come in many forms with dimensions ranging from the
' size of grain of salt to one that will occupy an entire room. The
. laser mcdium can be a gas, a liquid, an insulating crystal, or a
semlconductor, For optical fiber systems the laser iburces used
almost exclusively are semiconductor laser diodes. They are
simila.r to otherlasers, such as the conventionalsolid-state arid gas
lasers, in that the emittbd radiation has spatial and tempclral
' coherence, that is, the output radiation is highly rnonochromotic
and the light beam is very directional.
Despite their differences, the basic principle of operation is
the same for each t5rpe of laser. Laser action is the result of three
key processes: photon absorption, spontaneous emission and
stimulated emission. According to Planck's constant law, a
transition beiween two staes invoives the absorption (or) emission
of aphoton of energyfty = E2 - Ey
The electron can also be induced to make a downward txion
from the excited level to the ground state level by an external
stimulation. This emitted photon in phase with the incident
'photon,
and the resultant emission is Lnown as the stimulated
emission.
In thermal equilibrium the density of excited electron is very
small. Most photon incident on the systern will therefore be
absorbed, so that stimulated emission is esscntially negligible.
Stimulated emission will exceed absorptions only if the poputation
of eicited. states is greater than thai of the ground rt"ir. fnit
conduction is known as pbpulation inversion. In a semiconductor
laser, population iriversion is accomplished by injecting elecFons
into the material at the device contabts to fill the lower energy .

states of the conduction band.

l-2
|
I
I
#hyl2
.t
I
F.
K
Er E1 "l"
F.

I
6

',:

G
 ,.-unc..t^ilEr.
. .\
.,lY
B,
,: \

13 *
Optlcal Communlcatlon (EQ ;- 1 {OA
I

13. (a) (tt) A GaAs iqiectton Laser has an opfical cavity of lengtlt
250 ttm and width 100 Fh. At normal olDerating
temperahrre the gain factorB ts 2l x 10'3 A cm'S and tha
loss coefEcient a per cm is 10" Iletermine the threshotd

device, It iiay be'assumed that the cleaved mirrors are

uncoated and the current ls restricted to the optical
cavity. The refractive index of GaAs may be tdren 8s
3.6. (4)
Ans:
The Reflectivrty for normal insidence of a plane with on the
GaAs-air interface is
-
yL=y 2= y= (ffi+) 'z= (r#,,)'=0.32

The threshold current densitY is

' 1l**lrn!
Jsli=pL L rJ

10-rr lro++,r#rl
=_-
2l x ',==,, | - 250 x 10-* u'rz
LJ I

= 2.65 x 103 e.*-2

The threshold current 16 iq gtven by

Ittt = tg x area ofJhe optical cavity

2;,,65x 103 x 250 x 100 x 10'8

= 663mA.

(or)

13. O) (i) With the aid if diagrtms explatn the pcrmanent joining
schemcs of optical fiber. (10)
 ,.9
{,
'A"r

14

Ans:
Fusio
prepared fib6r ends. In this method, the fiber ends are first
prealigped and butted together. This is done either i'n a' gror>ved
fiber holder (or) under a microscope with micromanipulator$. The
butt joint is then heated with an electric arc (or) a laser pulse so
that the fiber ends are momentarily melted and bondod together.

The V-groove splice technique, the prepared fiber ends are

first butted iogethet in a V-shapa groove. They are then bonded
together witlr BII adhcsive (or) are held in place by means of a
cover plate. The V-channel can be either a grooved silicon,
plastic, ceramic, or mqtal substrate.
J--

Fibres to be spliced

V-groove
substrate

,..
.

::'r';

ffi
m

}
 *d Cqnmtnffir(EC-tlOal 15

.:, O) Gl) 100oC for a AIGa.A,s injection Laser wtth Tg = 160 K

13. Compare the threshold current densities at 32oC and
I

,l thc similar rado for an InGaAIP devicc wlth

"od
To=SSK'(O
:
. Ans:
The threshold current density

Itt " exp /r\

lU
' t|,4,luJ
I

For the AIGaAs device

til(3,zoe- n*, (i|lo) = 6.73
t,.
''
' .ld,ta\

'Hence the ratio of the current densities

tk(1004c) to.zs .
-,^
t,th(3zoq . o.73
---------=-;-=l.JJ

For the InGaAlP device

Iil'r3loc)cexp = 2s6.a7
H
I*r(loodC;,x exp*fffl = 881 .67

Hence the ratio of the current densities

ffi=ffi=3'44'
 .-r*dls "Lwf,s:
.Bt.
. -rF

l
16' B.E./B.Tech. Dpgros Exrmlnrtlon (Nov./ho. ZOO9) s

14. (a) (i) Describe the characteristic of Reslnnsivtty agelnst

wavelength for an ldeal silicon photodtode with the help
of its
expression. Explain the Dlrect and Indirect
. absorption in Si and Ge. (10)

The expression for quantum efficicncy data not involves

photon energy and therefore the responsivity R is often of more
. use wien characterizing the performance of a photodetector. It is
defined as:
I_
" R= f(,Aw-r)
o
... (1)

where I,
is thc ouQut photocurrent in ampcrs and Po is the
incident optical power in watts. The responsivity is a useful
parameter as it gives the transfer characteristic of the detector
' (i.e., pnorocurrent per unit incident optical power).

-Thr relationship for responsivity (Equation (l)) may_be

developed to include quantum efficioncy as follows. Considering
equation , thrir energy of a photon E = hf. Thusthe incident photon
rate rn may be written in terms of incident optical power and the
photon energy as:
P
o
'P= W
In equation, the electron rate is given by :

f =nf ... (3)

xQp

Substituting from Equatiotr, we obtain

NP
ro
tr= Tf ... (4)

F
 - l*sr-.-Sltr&G.
:,$
E,

Optlcd'Communtcttton (EG- f 4OA) 17c ,S

where e is the charge on an electron. Thus ftom Equation (1) the

responqivlt!'may be written as:
ra
qe
R= ... (5)
W
Equation (5) is a useful relationship fq respon;iviq which
may Ui Aevcloped a stage further to include the wavelength of the
incident light.
The frequency f of the incident photons is related to their
wavelength t and the velocity of tight in.air c, by:
tC
I=T .. . (6)

substr111lx,.,Hi,,,3ou** (5) a iinal tot the

responsivity is given bY:
".plession

R=g
nc
Responsibility

ldeal photodiode
0.88
\
Typical photodioda

1.0 Wbvalength (pm)

Flg. : Responsively against unvelength characterlstlc lor an ldeal

sigcon phoiodlode. The responslvlty ol a typlcal devlce ls algo shown
It may be noted that the responsivity is directly proportional
to the quantum efficiency at particular wavelengths.
a

The ideal responsivity against,wovelength characteristic for a

silicon photodiode with unit quantum efficiency is, illustrated in
 "rY{sr
.Qt.
,.lF

18 B.EJB.Tcch. Dcglrq Errmlnrtlon (Nov/pcg. rcOC)

Figure. Also shonrn is the typical responsivity of a practical silicon

device. I'

14. (a) (if) A,' fUotodlode has a quantumoe$ctency ?:7?% ,when

, photons of energy 1.5 x I0't' J are incident upon it.

operates and calculate the lncident opfical - Ix)wer

riqulred to obtaln a photocurrellof 2.2 pA when the
Photodiode ls operatlng as ebovg.
(O

The photon ener gY E = hf = hc/t

Therefore
h;,c / ), = 6,626 x 10-34 x 2.998 x rc8 t1.5 x 10-te

= 1.32 ptrt,
The photodiode is operating at 1 .3z.prtrl.
Responsivity (R) = qelhf = 0.73 x 1.6
= 0.77Anil.
Also ft = Iy'Po

= 2,2 x tO-6 tA.77

= 2.86;r\M

The ineidbnt optical power reluired is 2.86 PlW. t

,, ir (or)
l4.O) (i) Ilescrtbc the various sourccs of noise that affect the
perforrnance of Opdcal receiyers. Derlve expresslons
ior the S/N of APD receiYer. (S)
Ans:
The primary photoculrent ipn!)gonerated is,

0 =#,
a .'

l -,
br(r) = XP(r)
hy
 .:rf*-:*0.

t B.
s
Optlcrl' Communlcetlon (EC - 1 {02} 1g
The pnmary current cortsists of ,a dc value Io, which is the
average photocurrent'due to tho signal power, and a signal
component lp(t). nor pin photodiode the mean square sigmal
current
'.t

..t, = ir(t).
I
For a sinusoidally varying input signal of modulation index zc,
'a
the signal component <f >
'' .-]=4F.
. <j'fr)>
P" =f P =;-f.
ZP

The quantum noise current is gven by, :

.$, = ,O = 2e t, ** rgy
The bulk dack current ipp ariscs from e
-

,n

The total mean-square photo detection noise current

\
.frr,
can be writteil as, !
.,i,, =

= 2q (Ip + ID) u4'rgl1a + 2q IrB.

.z z 4KBT
.tfr=oT=TE ^
 i',,r,:.,.1. "- r*qQB.
.r, . +
'$.
20 . B.E./B,T€ch. Degrcs Examlnrilon (Nor./Drc. 2O0e) s

Thus the signal to noise ratio is given as,

s .Pr, M2
7rE 2q (lD + ID)Mz FW)B + zql ,B + (4KaTB/Rr)

14. O) (it) A good St APD (x . OJ) has a cepacltance of ! pF,

i negltgitle dark currEnt and ls operafing wlth a post
detection Bandwidt[of 50 MHz. VYhen the photocurrent
before Satn tg f0:7 I rind a tepperature ts I89C;
detcrmine the maximum S/It' lmprovement between
M ' I and M _' M--
_oD
assuming atl opera{ng condidons
aremalntalned (t)i-
Ans:
Maximum value of Load resistor Rl is grvcn by
ll
When M = 1, the SNR is grven by Equation
Thus the sNR when M = 1 is 9.0 dB.

When , - Mop and x ; 0.3, from Equation:

s ,,2+x=,'r;f PP
N=;;mg *ip 4KT

wherefa = 0andFn = 1 where Id = 0andFn = 1.

Hence, the shot noiscis:

p =2
2eBI^ x 1.602x 10-19 x 50 x 106 x l0-7
i , 4x1'38 1x10-23x}gl
rFt M2'2*
-op -
-

= t. 602 x 10-18A2
Optlcrl Cgmmunlcrtbn{EC :- | f02} 21 c

and thethermal noisgis: M oo= (5.255 x 103)0'o:t

0-23 x29l x 50 x lOn

:Er- =

The SNR at Mopmay be obtained from Equation:

M2 I2
I
I
1.263x 10-13 A2 $ =

It rnay be noted ma*=t the thermal

(41.54)2 x 10-14
-
 ..,x
R.
B.EJB.Tcch;Dcgrlc Exrmlhrtlon (Nov./Dro. eOOg) #

ls'

Ans:
(a) *
mffi ilHfrirffix[uxili,ffi:'[t'ts

The soliton solution to , the NLS equation holds to a

reasonable approximation only when individual pulso are well
sgpareted. to esure this, the soliton widthmust be a small fraction

This condition thus constraints thc achievable.bit rate, sitlce

there is limit on how narrow a soliton pulse can be generatcd.
If 7p is the width of the bit slot, then wc can relate the bit rate
B to the soliton half-max widih Tp by,
1- I
,=\=zsfr=m. 1,7627

ahg,physical clrplanqtion of the scparation rcquircments is

that the &CrfappingJails of closod spaccd solitos create onlincar ,

interactive fories bottneen thern, Thcse forces can be either

attractive (or) rcpulsive depending on the initial relative phase of
soliton
Optlcal Commuhlcatlon (EC-I 402)

,., a=texp(sfl t

"
The mutual interactive force between in-phase so that thus
iesults in periodic attraction, collapse and repulsion. The
interactio distance is,
L,r = L',--O
arsp = L---^rexp
paua (sn)

These tlpes of intoraction are ot desirable i a'soliton system,

sice they lcad tojittei in the soliton arrival times. One method for
avoiding this situation is to increase so. Since the interaction
between soliton dcpends on their spacing.

15. (a) (tl) Explatn the Layers of SONET wtth neat sketches. (8)
Ans: SONETLayers:
SONET has four optical interface layers, which are depicted
in the following figures and are as follows:
* Path
* Linc
t Section
t" Photonic

Mep$rrvlca rnd Prt

Ovrrhrad ln SFE

Mrp SPZ dc um
Ovcrherd into STS-N

irhp ST$N and Sgcion

Ovrrncrd inb hrLeo

Optlctl Cmvonion

TermlnrltrlUX AD0rlot Rogonrntor Rognnmtor ADM or TurtrlnalMUX "

Prth DCg Srqiln Sootbn DC8 Prlh

Trrminatim Liil Trnnhr0on Ttnnrnrtion Lfxr Ierminehon
Tcrmln*lon Tsmhrtlur
 - w.'%[$.
..*
B,
B,E./B.Tceh. O?erer Exrmlnrtlon (Nov,/Dro. AOO0) . $

I The path layer deals with tho transport of sorviccs between

thc path-tcrminating cquipment (PET) NEs. Thc.main functi<n of
llr p?m layor is* *lp ihc signals into a format required by, the
line laycr. ItS functions include reading, interpreting, .and
qo{ffinS tho POH forperformanec and ApS.

The STS PTD is an NE that multiplcxes or demultiplexs the

STS payldad. As an examplc, S,TS-l,irTE assembles 2g l.S4H.-
signels and insers POH to form a st.S+Mbps STS-I
YUry.D_Sl

l= Lf ni

Path

':
Ftg.:
' SOl\'ETTermlnsdonEqutpnont

The.linc layer deals with the transport of the path laycr.

I payload and itr orrcrtrcad'across thd physibal medium. The main
. function of the line layer is to provide synchronization and to
pcrform tnultipl,exing for the prath ,layer.. Its functions include
f protection gwitching, synchronization, .multiplcxing, line main-
tI tenan@f anderrormonitoring.
r
R.
lr'.
fls ,,
:
E
;.
i.t
 "'t- m,,
..r{
s.
Optical Communicatlon (EC-l 4O2l s

Line-Terminating Equipment (tTE)

. The line-termjnating equipmcnt (LTE) is the NE that
origrnates or rctminates the line signal.

section layqr..deals.with the.tr?nlPort gf an STS=f'l

'Thi fryc
;
across the" physical''mediuni. IS rirain functions are framing,
scrambling, erior monitoring, and section ma?ntenaneel'

The section'terminating equipment (STE)-,: the NE th.ut :"I

acb as a terminition device or a regenerator. The element is able
I

o"nnate' or termiuate its overhead'

illli:iflry,
t"
r
r'
I
,

liPhoto[leLay.ct. ,,
i
L
t-
,
The photonic'layer,mainly deals with the. transpo{ of bits
d.r*l the ptrysical-fiber medium. -t5 1ain-tun1tio1 is,.the
conversion Lettxreen STS-N and OC.N signals. Its functions .,';
'include
wavelength launching, pulse shapinE, and modulation of

15. O) (i) Describe the operations of unidirectional and Bidirec-

tional WDM. (8)
Ans: Uuidirectional WDM
'Unidirectional WDM systems rnul$p1ex
i.
a number of '
wavelengths for transmission in one direction on asingle fiber. For
exaniple, signals at various wavelengths in the C-band are
muttiplexed logether *or trErsmission over a single fibep. The
receiver receives multiplexed wavelengths.on a separate fiber.
Thi end-WDM device is responsible .for demultiplexing the
wavelengths and feeding thern to the appropriate receiyer. Figure
shows unidirectional:WDM. Unidirectional WDM system6 are
very common with cable plovidgrs who transmit multicast traffic
 B.E./B.Tech. Dqroe
Transmitter

':
A Multiplexing A Demultiplexing

A3 A3

A4 1 ,[-l lTl n+

Receiver -. Transmitter
A1

A2
.

A3
A3

A4f:J [In+
IFig;:
rE.'
:, , ,1
Bidirectional
'/ WDM i!

A bidirectional WDM system transmits and receives multiple

wavelengths over the sarne fiber.. For example, signals 0t various
wavelengths ,in the 1550-nm baSd are multiplexed together for
transmisSion over a single fiber., At the same timel: sopflfate
wavelengths in the 1550:rrrn band,are also received orel tle same
fiber. The end-.WDM device is responsible for multiplexing and
dgmultiplexing 'the wavelengths from and to their .respective
transmitters and receivers. Figure, shows a bidirectional WDM

Various tectrniquds are used to achieve full-duplex bi-diroc-

tional transmission over a single',fibor. Basically, the counter-
propagating signals on the same fiber have to be separated by
-suitable
using devices. Three methods are used to . achieve
:

| ,.,

'a

/
 s,
oprhrtcffinnrci@ e?

Transceiver

All A11

A10
{r0
A21 Fiber 1 421
,t!a

AZnt mAan
t-..

Transceivor Transcelwr
:411 All

A10
410

A21
A?1

A2n

Flg.:

Band.ScparadonMethod : , '
tn this method, the transmitted channcls aredividcd in two or
four,'gr6upir kndwn as su-bbands, travcling in oppositc ditoctions'
SuU-6mOi are ssparated and colnbined'Uy ofigt intorleavcrs
insorted:,ih line along the tranemission mcd-ir1m. fo pqvent thc
a_d-iacent bands rrom intertering udth each other alon;
thc
trausrnission fibegand ttr allow for easier band separation, a

ri.ctt.t gap kpown as a quar!'b*9 is lcft betrreen fhcm',Tfe

niea ,foithe guard-band-loads to inefficient.utilization of tho:
.r"il"Uf" specf,at bandvidtlr in bidirectiolal WD_!l systes and
tu"d;;"t"tty tirits the number of hansmissible chalnck.
' iypj;;lllt, inl numfir of wavelengths supportefl by the band'
separation methodis 32.
 .
ffiffii::'. ':,
'.F
s.

Interhavln*.f iltrr Method

The inteTleaving technique uses wavetcngth-interleaving
filters at- each end of the span. Interleaved channels arc usecl in
both directions of transmission. Even channels travel east to west,
whereas odd channels travel west to east. As a consequence,
channel spacing for wavelength's traveling in the same diiection
has to be dpubled. However, the interleaving filters have a high
insertion loss that contrib0tes to higher systemlosses.
Circulator Mothod
In this technique, the same wavelengths are transmitted in
both tlirections:of.propagation. To separarc transrnit and receive
direction at any node, optical circulators are used. A circulator is a
multi-port device that- allows signals to propagate in certain
.directions bascd on the port th.at the signal came from. The
r- circulator essentially acts rs rn isolator thai allows only unidiroc:
tional propagation.
15. (b) (ti) Draw apdr explsltr, the ppinclple of operatlons of an

gain. (t) .:
Ans: EDFAI Ttre active medium in an optical fiber amplifier
consists of a nominally l0 to 30 m length of'optical fiber thirt has
been lightly dop",6 with fofo-Earth .'it*"nts; such as ErSium,
Ytterbium.
The operating rcgion of these devices depends on thc host
materials and the doping elements. ?he moit popular doping
&?
ffi-
materials for long trautt tJlecommunication applic"tlon is a ritic.
fiber doped with'erwtri& is knolryn as EDFA.
ffi
":fr .
*i.
!*'

*
,$.,,

':!#
'ErDfum frbcr
{iP,
-,.ff*
Slgnal"in
iy-rF r$t '
.....,-. I
rS:,'
I 550 nm I

.','
,$n..

:
ff+
&
ffi,. CouPler

Ffg.: : An Erblum-doped llber emplifler

F
F
H
H
E,
F

k:?
 -: S'"*- G_
.F
s.

Optfica} Gixnmunlcstlon (EQ - 1 {q21'

a

980 nm

1480 nri

h
: r. I I Fig., ,

ffi
"x
ra*
jF
+
GatnFlatness i :

Since the populationrlcwls a1{r9 v+ious leyell wit\in a

band
rrr Aft".*t,'ttrOigatn of,:al EDFA becornes a function of'the
wavetcngth. Thtu In a WDM communicalion s-11teq' different
*bMlrfr.nnels rmdergo differont deqlees.of amplification. This is
.*tir"f iss.e, padCularly in WO!U, lYstems with cascadod
"
one way to improve the flames of the amplifrer gain profrel is
,o ,r." nouriOe el.tt fib"i instead of silica frber' doped with
;til Ei;;i. 3*t" ilpiin"tt are callcd ED FFAs, or- Erbium-
aop"a n'"oria" frber amplifiers. The flouride glags produces a
naiuratl flarrer garn spedhum compareC to sllica glass''H.owtJ:f'
there are arawUacrc to usinq flouride^gJass' The noisS
".fgi
' ;;;;ance of EDFFAsii pg:t nm than EDFAT' gtt reqsol is '

'- ihat they must be pumped t+eO and cannoi.S puT,ry1 ut

"i
'980 n.. t1 i. i Uecausl fluoride glass has an additional higher
Li"rgy t"""1 D4 above the E3levci, as shown.in Figure' with the
Af"i"n." in energies betwein thcse two levels cgrrTpoldin'.to
980 nin. This the 98p nm pump Porf,€r to- be absorbed for
".*t
 . :rFr -.ffi.r-r-
!
,9

. transitions from fhe E3 to E4level, which does not produce useful

, gain. This phenomenon is called excited state absorption.
' In addition to this dra,wback, fluoride frber itself is djfficult to
hhndle. It is bfittle" difficultto splice *ith conventional fiber, and
sysleptiA!9 !9 mgisture. Nev0rttieless, EDFFAs.are n9w commer-

, Another approach to fl4tten the EDFA gain'.is to use a filter , '

iqside the amplifier. The EDFA has a relatively high gain at1532
nm,:which.can.b,ereducedbyrninga,,i,rch-nIttir.in'that
w4vel6rrgth iegiorr inside the amplifies. Some of the. filters
described in Section 3.3 can be used for.tlris purpose. Long-period
I .fiber graqings and,dielectriq thin-film filtirs
"re
.u..uitiy. the

.l

.t

aa.,"i

ffi
1;*d

:;i,.
.. 11;*:

ii,i:. :.

.; l
t
 -:rqiq'',${ru#
i,s
g.

: ' B.E./B.Teeh.DEGREEEXAMINATION,APRIUMAYe0l0.
$eventh SemestBr

" EC1402,--OPTICAL COMMUNICATIO['I

(Regulation2Oo4)
(dommon to B.E. {Part-Tlme) Slxth Srmestsr' Regulatlon 2005)

t
Tlme : Three hours Maximum: lfi)Marks
AnswerALLQuegtlons
P.ARTA-(10X2=Mlrks) ,'
1. A multimode sep lndex flber wlth a core diameter of 80pm and
:a reladve lndex difierence ol1.5% ls operaffng at a wavelen'gth
of 0.t5pm;If the core refracf,ve indexls l;48.I)etermlne.
.
(a) normallded frequency of ltber.
O) thenumberofgddgdmodes.
' .\
Ans:
n1. = 1.48,1 = 0.85 pm' cotadiamcter = 80pm.

P fis.g'2
b)Mr=+='-)l=2873.

It is the maximum angle to the axis at which light may ?lttl

the fiber inordsr to be propagated. It is greater than ths critical
 -- tn.
's.

2 B.EJB.Tech. DcAreB Enmlnltlon (Aprll/Mry a0l O)

NumcrlcrtAperturel {. r .

' It i$ the-i;latioriship betw'ccn tho accepance angle'end the

refractive indices of thc threc mcdia involved namely corc,
cladding and' air. Tlio numorical apcrtlrrc is .a dimcnsioirless
' .quantity which is lcss than unity, wlrich values ranging from 0.14 to
0.50.

3. lVhat ls lntramodal dtsperston? How the'effect can be

'mlnlmlzed?

functio,n of wavelength. The increasing spectral width of optical

source will increase the intramodal dispersion. For exampte L$D
is having spectral ban{rnidth of 40 nm and it crnits wavelengths
from 830 to 870 nm butLASER diode has vcry narrol\, spectral
bandwidth having I or 2 nm. So obviously laser diode will'have
less intrarnodal di$persion ;

4. The optieat power launched lnto th" fiber is l0prw. The

,

transmlssion dlstance ts tOkrn. The optlcal power at the output

of llber is 2pw,
(a) calculate the stgnat attenuatlsn/unit km length. .
:

(b) calculate the over aII signal atteRuation.

Ans:
Signal attenuation per unit length
10' P, 10 logro
'.-i*tr1' ,a

^ =v.7.
,/
Overall signal attenuatioll B 0.7 x 1S a 7 dB. .

. 5. -'
State the three key process of laser action. D,efine them.
Ans:
" are certain conditioris to be fulfilled to achieve laser
"here
action. These are
 -,
!

t"!{

R.

Optlgal Co{nmunlqatlon (EC= 1 402)

a) The atomic system shoufd have population inversion. Popull-

, tion inversion meant thlstate 9f tle alom5 slste.T .t which
number of' atoms in the lower, or- giound slate, Population
inversion is achieved by pumping the atoms from lower state
'

b) There should be photons with proper e.nergy tor start the

- stimulated emission. '

c) These'should be an arrangement for multiple reflections to

,

t.

6. A lens coupled surface r emitting LED- Iaunches 190pw of

optical power into Iot$1o43 tfp index fibel whel a
forward current of "25 rnA is Ilowing thlough ther device.
Determine the over alt power convehsion efliciency when the
corresponding forward voltage across the diode is 1.5 V.
l,.ns:

O ptical P ow er
q= x I00Vo = = 0.506Vo
Electrical Power 25 x 10-5 .x 1.5 ,

I. f),efine guantum eflicfency,and responsivity of optical detector.

Ans: Quantum Efficiency:
ft ls defined as tha fraction of incident photons which are
absorb:d by the ptrgtoaetector and generate Jlectrons which are
collected aittre Aitector terminals.
','" t
Responsivity: ' t

It,is definedas the ratio of output photoculrent to the incident

optical power. .

ufreie ,,!p: oueut photo current, Po. incident optical power.

T .' ;F+.s.- +|FEXEFrr.
tA
- r"rt
s.
1. $
i4
I
l/May 2o1o)

: 8. Mention the begefits and drawbacks of avalanche p[otodlode.

.-:-
Carrier multiplication takes place in APD and as a result it
detects multiple electrons at.the output per incident photon. The
I
I
avalanche process in APD has:a sharp threshbld, which is sensitive
to ambieni tempera.ture and may require dynamic control of a
I
t
t

f.
I relafively high bias voltage.'APD with minimum gain however
I
I prolides about 15 db more receiver sensitivity than that achieled,'
-
F

with PIN photodiodc. Where,tlle applications, demand high

I
I

sensitivity silibon APDs are preferred choice.

': The disadvantage of APD is that it demands high birising
voltage and it is more noisy cortrpared to a PIN diode. Foi long
haul communications where received signal is weak, APD is

9.' lVhat is node parfitior noise? How Gan ft be reduced?

Mode partition noise is associated with the intensity fluctua-

tions in'the longitudinal rnodes of a laser diode. The side m'odes
are not sufficienily suppressed. This is dominant noise lw single
mode fibers, Mode partition noise becomes more pronounced for

Reduction:
i), The errors due to mode partition nroise can 6e reduced and
sometimes, eliminatedby seiing the bias point of the laser'
above threshold.
ii) The effect is reducBd by using a laser diode urith a few
longitudinal modes as possible. A single longitudinal mode

10. VYhat are the main parameters used for characterizing the
performance of optical amptifiers in a cornmunication sys[em?,
Ans:
i) Provide high gain

f,
 Optlarl Communlcrtlon (EC- f i[0e)

The maxwoll's equations' are

., .+
, _Ofi
VXE=ry

.-l

slnce 8,= pH .vx;E -_p#

 -r'rt
s.

{'+
..:
r
VxH trsinceD =eE r.'VxH =e At
: ;-.
tI
i V'D=0rV'B=0
t
-

.+
r,
i
I
.*

t-

i
!

r
; - Electric displacement density in c 7^2
,-
.+a
a B r Magnetic induction in Weberlmz '

r'
\,

C= ee
*r*o
r
i;

ffi
'ii*

B
#
fi:.i
.l'
?;:-
:!|i*

G
ilr
ry:
f.
F
4,.;.

k .a
* a(V x H)
k

L-
ki*
;e. Vx XE'= tt a, = -P.
H' (.*)
ffi
*
VxVxE =T tt € aZ E
sf

F
#;
h&.
ffi
dt* -:
d4f-_<.

H VxVxE =z v(v .El F vz1 = vz1 ('.' V'E E! 0)

ffi C
"&-
i- lt? a2 E
- vzr =-
r-1l*
:i ts.

af
b ,l
ffi v'E -l-
a(

H
t
r$'
E
L,
tp
#
 :---'wrr{Gr"
*
i^l{

Opttcal'Communleatlon (EC * I 4OA

\
-) .+
.4
Fpa(VXD) tt aN x € E)
VXVX H -il at
,' .
-t
@2
VxVx H ..t
t

I l;

'From Vector identity

..
-fn
-+

VxVxH=V(V.H)
vxVx;=?vzH
-l.t€ a2fi
-i2 H=
T tr 'f

peazH
'Yz fl
T.
*,

The wave equation shourd be convertgd to cylindrical

.oordi,i*r. ,yrt**: iht lil..tion oi plopaqafi:T' Th: magnetic
light rav
propagating-it i""gt, the fiber axis lryins electric
.ld
eac,h other and also,
field components which is perpendicular P
l

The fields components are given as

-'
B7#HoeJ9t_p4=:pHogi{*t-fu),i'*
^br
H = Hg (,r, Q) sif*t - frz) where B = +. .

To solVe Maxwell's equation let us take'

I
 B.E./B.Tech. tlegree Eximlnallon (Aprll/May 20f O) a;

t' '{ ad
I
o'l :.r

1la a
..

v E=:ilE- w irl
A-I =p
at
AH

oEo ,

\'
1", "rl l'

--#Hs si{*t- Fz) a, p Ho si!'t - Pz) 'lu

t

= i'

It
'
a

= - w Hg eit*t'-
u) fiz)
= - jo P,H. *.

'.. ''
F

ar l rE, o(nEJl
vxE=TL@=tr)
,l

[rf, ,Erl o, [aE

lr I
aE1
rl
lar aa)
l--

V X E'=i j,p.pH. ,

The decomposition of magnetic field is,

i'

'j. H = Hrar* HOaq + Hzaz

'

V rx e: !$cla(Flr a; + Hq a6 * Hzaz)
d("r)l
i,tHr = T fuu,
1 :

lW--ru)
l lan -l = -iFEo'
=;L@ +irfto]
.-

dB,l
-i*trH*:
''
l; - *l
\\J
v'
I

i*pHy = ifr E, +
I " a(E )l t
 :q.F
.(,t{

.,q.
Opthal Gommunlertlon (EC : I 40el *
Similarly,

.1t", *!
o: ;

vxH=ll* * * =:"e E

il.
,l
* !'

[Er ar + EO a6 &azJ
(
= JCIG
a
t

iueE,=lF-i tB"i
aHz

JCI z
€ E,'= frr# -:'l
lrL*dr @.-j

L,
.I''FE,
I + ;- 1
dE
,
"o- i*n i*tr- dr
=

1 dH,T (iFlz Ei . ifr d Fn

u*r -7W
tw 1*i ' jrp 0r

-a

where ,2 st,cnl*=q2=t?-*
-1 I irp
oHo ipa Erl
Erqo
=*;
L' :&-*-T1 J
t'

-i I fr
dE" ttotlH rl
fi, . =.-*
E.
.. :.i: --_,,
9qn L?E---T-J
,
 -.-.q:,,%I.E
s
.r,lt

10 B.E,/B.Tech; Degree Examination (AprillMay 2ot 0)

. f BaH, aEl
zl
t
Hr'=i aC
i
rll
a

H r -t
a0

aE-
J
I

Hr=iF AHa0 z
I @€Trlzl

1
aHl
rl
r ao) =jroe E_
z

p ozHr. a)€ ,2 H, .- jP Hz , ,e
-##+iiT.rtffi.?W=iu az azE*
€Ez

azE z oE,
nI
a"r
J L*+1
,2 ooz r f
+ q"E, =0 z

azH
'zlzlz .taH . a2H
-p+;a, +
?W + qZH, = Q.'

Modal equations for step iudex fiber:

We know that for a step index fiber, the refractive index
profile can be formally expressed as
n(r)*= ,r<a
,f7a
The mode fields ca4be obtained by solving the wave equation
separately inside the core and cladding regions and matching the
fields at the boundaries.'The modal equafions are given by
oE, 1 azL, +qZEz:O
az E,
*- 2
+:r1 dr
^- {'-?W
dr

Lrz Hz 1 oH, 1
+
az H,
qzHz:o
i+7 a, ?@+
I

l
20
 r-qGg. *{-
#.

s
.
j,B

11
Optieal Communication (EC:l 402)

where qz =

The guided rnodes should satisfy the condition

konz< B . kon!.

Due to circular symmetr ; variation with Q is periodic

and so
in each region
F' . a,
= AF(rf
it o - fiz)
.LE'- 'it*t
H, - BG(r1
- frz)
"i'a 'i('t
where \/=0, 1,2,...
with this the modal equation become

F-0
l

#- 1#.[,, -f,cG='0 ions depends on the

The solutions in the ocre and cladding regio
' r!
slgn- or q and aro given by the bessel functior
ons of order v and
*oatn.d br*sel functiort of the second ordor v re
respectivelY.

G(ul = Clov (ur/a), r( a

DK/\ v
a
 .-s(F..r@t$g.
f -'lt
B.

72 B.E./B.Tech. Degree Exarnlnatlon (Aprll/May 2Ot O)

EO and Ez are same at interface

HA and Hz qre same at interface

tr

Applying these ccinditions to eliminate the arbitrary constants

A, B, C, D.
- K ,(an). 0
v\
lau
a\

rK-.(ua)
#Ku('" ) --@ V\

0 i
.
i;,;,- K r(Ala
.'i
.
)
- ia,9
#Kr(4, a) trv Kr(ara)
?a
=
0.

ation.
(Fr\2 I t 1\2
lzj [r+ ,2 I I

Iu(ua)
where I=
v uJr(u a )
.,
Ku@a)
)*
m,at'. '.r

(or)
11. (b) (i) A typical refractive index difference.for an optical fiber
designated for long distance transmission is I Vo. Deter.
rnine the NA and the solid acceptance angle in air for the
fiber when the core index is 1.46. Calculate the critical
angle at the core-cladding interface with in the fiber. (6)
 .;,n
s.

13
Otrbal Communlcatlon (EC-l 402)
{-
Ans: nlI = 7.46
A =l%o =0.01
Numerical Apertore = sin Q= n1 ffi
= 1.46{2 x 0.01

='Q.206
.0=?
NA=sin0= 0.206

Q = 11.9o = Critical angle

Acceptance angle,
fi =n*
= 444.65.

11. O) (ii) Derive an expression for nurnerical apeerture of a

graded index fiber structure' (10) t:

Ans:
given bY,
The index profile variation of a graded index fiber is

n(r) = [1r - La,(r t o)"1r'

2 for0srsa
2^)L'2 = nt(1 A) = n2 for n z a"
b,,
The index profile will obey the power law relationship
ingraded index nUr* The refractive index is not constant
in
fronr- a
graded index fiber but it decreTes with radial. distance
maximum value of n, of the axis to a constant value'oz
in the
cladding
where e = 1, the profile is triangular

ct a m, the profile is constant anditwitl become astepindex fiber.

\
 B.E./B,Tech. Degree Examlnatton (Aprlll"May 2Ol O)
s

u=inf

Fig. :tndex prolile variation forgraded indexfiber

12. (a) lVrite notes on

(i) Absorption. (5)
(ii) Scattering losses. (5)
(iii) Bending losses occurrlng in opticatr fibers. (O
Ansl i) Absorption:
Absprption is caused by three me-chanism
i) Absorption by atomic dpfects in the fiber materials.
ii) trntrinsic absorption by the basic constituent atoms of the fiber
material.
iii)Extrinsicabsorptionbyimpurityatomsinthefibermaterial.

Absorption by atornic defects:

These type of absorption is caused by atomic defects like
missing motliules, clustJr'of atoms, impeifection of the atomic
structure of the fibprmatcrial. This type of absorption having very !

smallvalue,thevalueismostlynegligiblewhencomparingwith
intrinsic and extrinsic absorplion. ft e totat radiation energy !
received by the fiber is r*pt"ised in units of rad (Si), which is a i;
measure of radiation absorbed by a silicoll.
l rad(s;) = o.ot I/Kg.
 'iffi::3qh:

:' Optlcal Gommunlcatlon (EC-t402)

',n -
t5 ..s.

s
. When the radiation level increases, attenuation- will' get '
increase. At the time of radiation, iadiation,particlep will be
produced. The particles or rays.such as electrons, neturons,
--
ga*tn" t"y* are affect thestructure of fiber.
Intrinsic Absorption:
The intrinsic absorption having two types
i) Intrinsic absorption due to ultraviolet fail;
ii) Intrinsic absorption due to IR absorption fail.
The optical communication wavelength+angb in terms of pm
is 0.8pm ti t.lpm.In silica fibers, intrinsic absorption will occur
above 1.5 pm. The photons of light energy are converted into
random mechanical vibration infrared absorption. Maximum IR
peak value at 0.Q4m and minimum peak value at3.2,3.8 and 4.4
, i^. The emperical expression for the IR absorption for
GuOz - SiOzglass is grven bY,

a IR: 7.87 x1011 x (ry-) . '

" \^l '

Intrinsic absorption also depends up.on the electronic absorp-

tion bands in the UV region. UV abSorption dec4ys exponintially
with increising wavelength at 0.8pm the ulffa violet absorption
having the value of 0.3db/km. Thus in the case of silica frber, it s'ill
acf as a transmission window from lpnr to X.61,m.
Extrinsic Absorption by Impurity Atoms:
Extrinsic Absorption is due to fansitiol tetal.lgns suctr as
iron, chromium, coPper, manganese, nickel. This type 9f
absorpqion is more pronounce in direct melt methods,-b.ecause.in
, that tygie Of fabricationrnethod, the dopants are adde{ directlyto
the siiica. In the case of Vapour Axial Depositioh (VAD) the
impurity lovel ranges from 1 to 5 parts per billion. The transition
. metal ions produce loss at,t = 0.8pm.
Impurity absorption also results frorn OH ions. These OH
impurity results from the oxyhydla!9n _fPT-u Le$ for the
hydrolysis reaction'of the SiCla,GeCl4,POC$. This type of
UE.;

B.E./B.Tech. Degree Examination (Aprll/May 2ol 0)

r.*
16 B.

{ absorption is reduced by reducing the water content in the fiber

;r

\
Scattering losses having two types
i) Lisrar scatiering loss ii) Non-linear Scattering loss
a. Rayleigh scattering' a: Stimulate Brillouin Scattering
b. Mie scatting n. Stimulated Raman Scattering.

Linear Scattering Losses: F

Linear scattering transfers linearly the. optical powqr in one

propagating mode to a different mode. This losses will occurs as a
-mode
iiuty or as radiation mode. Scattering loss yll be,more in
mutiimode fibers due to higher dopand concentration and greater
composition al fluctuations .
a) Rayleigh Scattering:
This loss occurs in the UV region. Its tail extends upto IR
region. It arises from the microscopic inhom!8e1etics.Ptt:tnt in
"

th[ material of fiber. In homogeneities may arise froT th? density

fl uctuations, refl ective fluctuatioris an d composition al vari ations .

cscat
^
=
#
n8
,,

f $, rcT, m-!"
a a , , f L a ,, a

wtrere, r + Refractive index of silica.

p* Photoelastic co-efficient of silica. .

F C* IsothermalcompressibilitY
Tp .7 Fictiye temgrerature at which solidification of glass

The transmission loss due to Rayleigh scattering

where, L * Length of fiber.

t
I

f
 .w.r:q[;.
- Optlcal Conmunlcallon 1qC.frce1 17 "S'
.'.'. .:' .

At high wavelength Rayleigh Scattering LoSs will be reduced.

It is an elastic scattgiing becauSe there is no change in frequency.
b) Mie Scattering:
, Mie Scattering arises due to the imperfect structure of the
wave guide irreguiarities in the core'cladding interface, corq;clad-
ding refractive indbx difference along-the fiber and diameter
fluctuations.rBy achieving high relative refractive index difference
anddoingperfectfabricationMiescatteringgetreduced.
Non-Linear Scattering Losses:
Non-Linear Scattering Losses will occur at high Power
Levels. When the transferring of one mode to other mode taking r
' place the output will be forward or ievorse direction. This is
. inelastic scattering due to shift in the frequency when the
refractive index of the medium depends on the optical intensity of
. tlie signal, then these non-linear scatterings are occurgd
a) Stimulated Brillouin Scattering:
It is a loss mechanism due to thermal molecular vibrations
within the fiber. This type of seattered light contains uppr side
band and lower side band along with the incident light frequency.
Thethresholdopticalp.owerform,BrillouinScatteringispropor-
tional to
&*o,
whered-Fibercorediameter , '

I * Operating wavelength
ant', * Brillouin Scattering Loss Co- efficient.

Here, an incidgntligbt will procedure scattered photon as well

as a photon ofacoustic frequencY.

b) Stimulated Raman Scattering:

The scattered lighr. consist of a scattered phofon and a high
frequency optical photon. In contrast to Brillouin scattering,
Ramah scattering is having three orders of high magnitude. The
threshold optical pwoer for Raman Scatteringis proportional to
 .=<8".*i_
S.

18 B.E./B.Tech. Degree Examlnatlon (April/May 201 0)

*
...*

.&
.n
trLa R.

. where 4 - Core diameter

I - Operating wavelength
aR -f Raman scattering loss co-efficient.

. iit) Bendinglosses Occuring in Ophcal Fibers:

Whenever the $nds and curves will be in the path of optical
ray radiative losses will occur. There are two types of bending
losses,

ii) Microscopic bending losses.

Macroscopic
a
Bending Losses (or) Large Radius Losses:
These.occur when the radius of curvature of bend is greater
than the fiber diamlter. When the radius of curvature of bend
decreases (or) curvature of fiber increases, the loss increases
exponentially upto a critical radius of curvature.
Formultimode fibor that critical radius of curva.ture of bend

3n:?

.'Rc'=@fi
The attenuatiol co-efficientby the macrobends

ob= A exp (- BRc).

Fqr single m oo: fibe

2.4!1
RCE
(nt - fl:z )3/2

: wher"lc = Cut offwavelength

?n anL lX _* Z?ra(NA)
=- 2.405 z.4os
 19
Optical Communlcation (EC''-I 402)

Gladding

Power loss through radiation

Field distribution

Bent flber

Fig.
Minimizationofthistypeoflossesisdoneby
i) Fibers with large relative refractive index difference.
ii) Operating at the shortest wavelength possible.
Microbending Mode Coupling:Losses):
It introduces from the fiber when it is incorporated into
cables. This type of bending introduces stight surface imperfec-
tions which can cause mode coupling between adjacent modes or
coupling of energy between the guided To{"s and the -leaky
*oi.* in the fiberwhich:inturn creates a yadiativ.l Ioss.-ThQ losses
due to non-uniform pressure during cablinq t! rdferred 1s calling.
losses. The loss is depends on the fiber deformation,.length of
fiber and the optical- power distribution. Microbenling losses
proportional to if,. number of modes proqagating through the
to wavelength.
itUrt and inversely proportional
(or)
,if

t2. (b) (i) Exptain the design optimization characteristics of single

Ans:
Single mode fibers are used in the telecommunication
in
applications ryd oPtical networks, the disperslon Produced
single mode flber must be reduce. This can be achieved bY the
 .=rar-.''Yl ,

20 B.E.lB.Tech. Degree Examination (Aprll/May ZOI O) s

r,lt

$
i:
!:t
tr
Jr.' suitable design of refractive index profile and, making the addition
t
of material and waveguide dispersion is equal to zero, _
i There are three ways of shaping the refractive index profile i)-
1300 nm optimized fiber, ii) Dispersion shifted fiter, iiii
Dispersion flattened fibor, iv) Large-effective core area fibers.
:

i) 1300 nm optimized fiber:

At 1300 nm; single mode fiber is having two types:
a) Matched cladding rpfractive indexprofile;
b) Dispersed cladding refractive index profile.
Matche-d cladding fibers having uniform index profie
throughout the cladding. In this type mode filed diameters are 9.5
pm and relative refractive index difference is about 0.35 Vo.In the
dispersed cladding fibers, the cladding region surrounding the
core has a lower refractive index than the outer cladding region.
Thus its refractive index is in the form of a depressed manner
surrounding with high refractive index iegions. ftris is the type in
which the waveguide dispersion value will be zero at 1 .3 ttm or

Even though there is no waveguide dispersion at 1 .3 pm, there

wiil be some Rnite loss. Ther.foi. now ; days the single mode
fibers are designed such that zero'disporsion at 1.55 #m with a
minimurn loss. At 1.55 pm, the material dispersion having positive
and large value, the same time waveguide dispersion ir f,uring
negative, and small.value . '

To rnake ttie tot;l dispersion value is equal to zera, the

relative refractive index diffrent is slightly increased by adding
more G"Oz in core, a triangular refractive index profile can be
designed. Thus the dispersion shifted fibers have minimum loss
and zero dispersion at 1.55 Fm.

I.
B
\?
 - .r.lt
R.
Optlcal Communlcatlon (EC-I 402) s

iii) Dispersion Flattened Fibersi

This tlpe frbers having minimum dispersion over a range of
v,ravelengtl; from 1.31nm to 1.551nm. Such that zero dispersion
points tie at 1.3 pm ald 1.55pm. These fibers can also be used for
iavetength division multiplexing. The refractive index profile is
modified. So the dispersion value is shifted to zetoi
There is two type -

a) Double clad or al Profiel.

b) Quadrupole clad profile.
ar profile type with steepr maximum refractive at the center of
the'.oi, and ahe quadrupole cladd is the improved verlion of ar
profile. All these refractive index profile designs are making zero
dirp.rsion at 1.55 pm by adjusting the magnitude of waveguide

iv) Large Effective Core Area Fibers:

Large core area is the need to reduce the effects of fiber
non-linearity. TWo examples of the indpx profile is
a) LarBe area dispersion shifted.
b) Large area dispersion shifted.
l2.O) (ii) A multimode graded index fiber exhibits total pulse
broadening'of 0.1,,as over a distance of 15 km. Estirnate
(1) the maximtun possible band$dth on the link assuming
no inter symbol interference. (6)
(2) pulse dispersion/unit length.
(3) bandwidth-Iength product for the fiber.
Ans:

1) Bopt =

r 0.1 x 10-6 6.67nSkm-t

T=T=

h
22 B.E./B.Tech. Degree F*amlnatlon (Aprll VMaY 201'0)

3) Bandwidth length Product

= 5 MHz x 15 Km =75 MHz Km'

13. (a) I)escribe with a neat diagram variorrs LED structures
and explain their characteristics. (1O
Ansl
Fiber Bo
Circular Metalization

r Double leather
Confinment jiunction layer
layers

Circular metal contact

Fig. : schematic of a high radiance surface emitting LED

Strip
contact Active area

Metalization
Light guiding
layei's ,f Si02 isolation laYer

\i lDouble leattrer
J junction layer
Substrate \

Metalization
(for electric Heat &
contact) ink
Edge Emitter
lncoherent
optieal output beam

Fig. : Edge emitter

 _.._
s,a-..rgtrHi!.
..9
B,

Optlcal Communicatlon (EC-t 402)'

The requirements fog an LED to be used for fiber transmis-

Sion are [a] high radiance output [b] high quantity efficiency and
[cJ fast emission.response time. LEDs radiance or brightness
is a
measure itr watts of the optical power radiated into a unit solid
. at'rgle per unit area of the emitting surface. Emission re$ponse
time can be reckoned as the time delay between the application of
a current. pulse and the onset of optical emissiorr. The quantum
efficiency'is related to the ftaction of the injected electron hole
pairs that recornbine radiatively. LED structure has to be'
designed to provide a means of.confining the charge carriers and
the stimulated optical emission to the active region of the pn
junction where radiative reeombination takes place. This enables,
to achieve high iadiance and high quantum efficiency. With this
- approach high quantum'efficiency is achieved. Optical confine-
mint is yery important fo1 preventing absorption of the emitted
radiaiton of the material strroundingthe pn junction. 'l
LED configuration such as homojunction and single and
double hetbrojunction have been identified as potential can-
didates of achieve carrier and optical confinement. This is double
heterostructure or heterojunction device. It is possible for both
carriers and the optical field in confinement in the central active
layer, dup to this sandwich structure of differently compould a!191
layers. Both high efficiency and high radiance isrdue.lo this dual
confinement. Optical absorption in the 'active region, carrier
recombination at the hetero structure interfaces doping con-
centrations of the active layer, injection carrier density and active
layer thickness are the various other Parameters influencing the
device performancc. Thu"effects of these parameters will be
discussed below,
Trvo basic LEDs configurations used in fiber optics are
Surfacg emitters and edge.emitters. In the surface emitter, the'
. plane of active I.ED region'is oriented perpendicularly to the axil
of the fiber as shown in figure. tn this configuration a well is etched
through the substrate of the device, into which a fiber is then
cemented in order to accept the emitted light. In practical surface
 ',F
B.

24 B.E./B.Tech. Degree Examination (April/May 201 O)

 \

j.*

B.

Optlcal Cornmunicatton (EC-l 4OA)

Coupling Bf,ficiency (f)

It is defined as a measure of,the amount 9f optical power
ernitted from a sourclthat ca1 beroupled into a fiber

rt= ps
-'
wtrere ps - power couplt: fib1r
:"ro
nitted from light source. F

Pigtail (or) Flylead:

Many source suppliers offer devices with a short length of
optical fiber already attach,ed. in a optimum power-.?upling
configuration. This section of fiber is generally referred to as
flylead or pigtail.

Power Coupling Calculations: ,

i) Power coupled into step index fiber from surface emitting LED.
When r, >a
r^h

= ?o2*, Bs n! L^.

Here rs - radius of the emitting are anr! when rs

F)

( {t.
,^Y

Plnn, step = *r? * Ao n? L (oi) n2 Bs(N-4 )z *

'of the sourc" irp, = nz 4, Bg.
i

Here the radiance Bg is the emitted power per unit area of the
emitting surface per unit solid angle converted by hemisphere per
unit angular speed of radiations.
 :_ri;p_ *,:@f,lQG.

"*
E,
2e

:Ps
:. Thrs86,=T /,
2n x n6.f cos 0 d cos 0
0

it) ower coupled into graded index fibet from the surfaee
P.

cmitting LED
r"
Pr-Eb, graded = 2'? Bo{*," o ntrl' dr
P

,L

J,

The coupled po-wer from the Laser or'LED can be written as

''

where X - LED or Lasgr diode spatial distribution factor

= 2to3 for edge emitting LED

.:
This is true only when rs < a. ' , (

.\
r .l
Fresnel reflection coefficient
- :

' r r.2 .' "

"
- n\-
r

ln.
R= ltt., ,-1 .
f..\.t l\*al

tt
21
 *,:.E@EE
==qat*
\
s
'i.*

Optlcal Communicatlon (EQ,}- 1 {02} {k

3

Coupling Efliciency: !

pi Power into fiber

n,g =-=
ps

/Pemitted\
CouptingLossindB = 10 log
tog lffil. '
le*'el'dJ
i) When the source emitting $urface and fiber entrance end are
sepprated by a distance 'd', then the coupling efficiency for
step index fiber,

?st=l@i Ll
'
lr +
ii) For lateral displ...,r\rnt 'd' the .o,ipting efficiency'

2 --
d
?st = It co$
iH fi

Ilt, For graded index fibe r, in the case of longirudinal displace-

aaa\"

ment'd',

_m,6)
I

r
2l
,I

?st =
-1G {d\
"'lcos \m) ('€']
L
..

But if d < 0.4a

,l ,
:

8d
1
?ct = 1
3wfl'
-e:ri

.l

Power Launching Ys Wavetength:

Power Launched i,nto sipgle mode fibel {oes not deqe.nd qn
wavelength. Power launched into graded index'fiber or multimode
fiber depend gn wavefength.
 e8 B.EJB.Tech. Degrso Examlnatton. (Aprll/May 201'0\' *

Number of modes can propagate through the riber,,

(?fian'\' 1

I
i-..": '
rrF.l.-."
,
M=#t---l
I a
L-T|ttf, . I

a.

tf
Do
... Radiated mode at each mode =
#" Blg tt

r13,. (b) (ii)'With examples describe the possible lensing schemes

If the emitting area of the source is stnaller than the core area,
a miniature lens tirry be placed between the source and the fiber to
improve the power eoupling efficiency. The function of microlens
is to magnify the emitting area of the source to match exacily ttie
core area orine fiber end facq.
The different lensing schemes are shown to improve the
coupling efficiency.
The figure (a) shows that the emitting area is coupled with
rounded end fiber. Here the whole radiation is incident fully on
thle fiber end surface.
The spherical surface of LED and rounded end fiber is shown
in figure (b). Here the focussed radiati0h ftom the source is
focuJsed fuliy on the3ore are. The figure (c) shown the imaging
sphere whict, ..iuptesioth source anci itrr fiber. Here the divergeci
radiation from the source is fully focussed on to the core Broo.
The microlens whose diameter is equal to diameter of the
surface area is shown in figure. When thC microlens is in contact
with source and fiber such that the refractive index of lens material
' is equal to refractive index of the core, maximpm coppling occurs.
meth,pdis maximum efficigncy of couplingiS obtained. .

The coupling of sources with taper ended fiber it :.t ?Y i1

figure (f). If it" *iOtt of the taper ended fiber is equal ts'width,of
ng surface and the $ourc€ is in eontact with taper ended

v
LED

(a)
' ,,).
Active area

Round&end!i&r
Yu'rn

(b).Non-irnagi ng rnicro sPhere

-.

E
(c) Imaging sPhere (d) Cylindrical lens

(e) Sphericai-surface LED (t) Taper-ended fiber

':}

14.(a) photodiode.
(i) Expliin thg detgction principle of avalanche
ouiain the relationship between responsivity and quan-
- turn effrciency. (10) :

Ais:
In contrast to PIN ptrolodiode, which gain of unity
lt:It.nt
1000 but it can go up to
an ApD has a better gain. It can be around
 -ra{d-. i.EWh1,
- ..q
B.

i,
.{
B.E./B.Tech. Degree B(amlnaliofi {Apillliflay 2ol o) ;f

several 100 as'typic'al value. This is possible mainly because of

avalanche multipti'i:ation, because of high intersity field. Also if
the kinetic energy of a ca-ier is greater than the band gap energy
of the valence electorns, the eollision can free a bound electron.
The free electron and hsle so created themselves acquire enough
kinetic energy to cause ftrliler impact ionjzation, The re*ult is an
avalanche, with the number of free carrier's gr.owing exponential-
ly as the process continues.

, The RdPD'cqnslsts of p* - i - p -,u* layqrs, TheTllayer

is basically intrinsic matcrial and in advertently has some p doping
of imperfect purification. When the diode is reverse
P*olt mo,.s1
,bias9d, of applied vo age ap,pgars
lhe the pn' 4cross
junction because of negligibly smai{: ptoto current with the
'increase in the bias voltage, peak Clectric field at the jurrction and
the width of the depletion rqglon increases. At a certain voltage
level, thO electric field is about llVo less than the avalanche brbak
down limitand depletion layer juSt reaches through to the nearly-
in intrinsic z region. The RAPD is generally.operated in the fully
depleted'mode. Photons enter-the device through'the p* region
and are absorbed'by'the high'resistively intrinJic P-tlrec Layer ,

wirere electron-hole pairs *"g"nr."ted. The retativ6ly weak

clectric field in this region then separates the carrier causing the
el€ctroms and lioles to d*ft into the'high electric field 5egrbn rihere
avalanche multiplication occurs ,"r,lting in current fain,of the
device.
Relationship between responsivity and Quantum efficlency
One measure ofphoto diode sensifivity is the responsivity R6
which is defined as the ratio of detector curren! to the input optical
power,level. It is normally'expressed as AAV oi pAtpW.
Responsivity is given by' :

e (or)
q
,^-\ )'4 e ri, / pm A
B0 =,7 =
,r*w w'
r E
 ?E.dB'_ yJ%E
't
s.
'3{

- Opthrl Gommunlcatlon (EG-1402)

'Responsivity tunction l:^and ghoJo.diode material'
if l 9f its respOnsivity,
sensitivity of an App is deereased by

RoM
 32 B. E.lB.Tech. Degree A(amlnatlon (Aprll/May 201 0)

' l4.O) (t) Discuss the varlotu noise sources and disturbances ihat

' Ans:
The different noise sources are discussed below:

i r), ShotNolse(Quantum):
The principal noises associated with ghoto detectors having
no inte;nal gain are qqantum noise, darlc ctirrent noise gener?ted
. in the bUIk material of the photodiode and surface leakage culrent
noise.Theshotnoiseorquantumari9esfromthestatisticalnature
' of the production and collection of photo electrons when an
optical signal is incident on a photo detector. These statistics
follows a Poissoq procass-. The quantum noise eurrent has a mean
square value in a bandwidth B.B is proportional to the average
: . values of the photo currcnt Ip, '' '
P t^ a*iz
e = z^qp rr,r,1

: where F(uy.is a noise figure assoeiated with the random nature of

the iivatanche process. From experimeritai results. F\M\'= l'f
where X depends on the materiai. For pin photo diodes M and
F[M] areunity. r '. , ,
'

!'
' b) Dark Current:
This is the current that continu'es to ,flow. through the bias- , '

circuit of the device when no light is falling on thq photodiode.

This is a combination of$ulk.and surface currents. T!: bulk
-F current. ipa. aris-es from elcctrons and or' holes, which are'.r
'
thermally generated in the pn iunction of the photodiode.
; Therefoie ttris is multiplied by the avalanche gain mechanism.
.Meansquare*,'T}.;.].1';,|,o*,,iWn,.,
F where Io is the ptimary unniultiplied detectorbulk dark current.

.t
Optlg.tl Communieatlon (EC- 1 4o2l

ments,.
 ' -.-- a.l- -'-: r.
- : --"t
- .: ,T.*o#'-r. q.
l
"ls
t
S.

34 B. E./ B.Tech. Deg ree Ettaniinatlon (Aprlllttltay PI O)

diagram the design of

' o \ao
.- t . i

14. {bxit) Explain recdiver

^yth leat
preamplifiers. (E)

The three basic structures of preamplifier are Iow irnpedance,

high impe dan c6 and transimpe d ance pre arnplifi er .-A pre ar.nplifi er
should satisfy the following requirements such as low noise level,
tiigh bandwiAtn, high {namic range, high sensitivity and high
galn,
dance Preamplifier

&
P-i-N
Photo diode

It is a CE amplifier with the p - i - n photodiode in figure. It has

,

a' low input irnpedance and thereforp it gives a rnoderate

'

bandwidth operaiion without the need for equ{izatiol..Therefore

for'low noise operation one should select wheth.er high impedance
preamplifierortransimpedancepreampIifie.r.
High Impedance AmPlifier
It offers lowest noise level and hence the highest detection
sensitivity. Ii we make the R1 large for either the FET or bipolar
front end, then the amplifier i*puiimpedancg $rilt be high. This is
the reason for considering the high- impedance at the front end the
large R1 reduces the receiver bandwidth. The trig! lfp:dance
!, amptifier operates above 3.:dB banwidth, but amplifies higher
t,
frequencies-much less"than lower ones. Equalizers are placed in
i'- -.

h
 .tl:j{;:-1,::l
.-,..,
!.. : ;
-:+s" .a. p.
.
t

S.
"!{

'a

OSleal Communleatlon (EC;t 402)

the receivcr somcwhere after the.preamplifier. It is used to

improve the noise characteristics of high impedance front ends:
The high impedance front end does not have a wide dynamic
range beauso of the large load resistance this problem is solved by
the transimpetlance amplifi er.
Vcc

&
P.i.N
Photo diode

Fig.' Hlgh lmpedance FETcommon source preampitfier

Fig. : Tra nsimpedanee arnplifier

The transimpedance amplifier operator over a wide dynamic
f,rnge. High vaule. of feedback resistances will eliminate the
ivCr sensitivity degradation. The desired bandwidth value can
achieved by increasing the open loop gain of amplifier; No
 -.
.. I : - .
.

.F

,'. E

B. E./B.Tec h. Dcgrla Exrm|nrtlpn (Aprlllffiry 201 0)

equalizing circuit are requirgd. High impedance FET p-toamplifier

can Ue usla for up I Gbps due to its low noise levol. Abde l 6bps
transimpcdancc'amptitiers are uqofu] duc to its widc dynamic

-- :'' (i)-' what rye the key system requirements ..and charac'
15. (a)
t rlrd.rrcqulredllanatystugapolnttopolntllnk? (O
Ans:
,

tt ir it r qimplgst transmission:linkhaving alransmittef on one

end aqd receivir on the. othet cnd. The-reryat3rs yay bc opto
electronic or'optichl repeaters. Thg maiol dosig4 factor of the
l
,
systern is repeaier spacing, If the repeaterspacing incroases, the

' For analysing'a link, ihe key'sy(1rm requi1gle3ts are the

mCxiinum transmission distaricc, the datA rate' and B ER r

Issatisff,thesc iequirCments tlc syste* thould be designed

based onthe ttmponents andtheif chalactetiltics: , :

l) Multimode or singlo modc fiber.

I. a. Core radius
b. Rcfractive index Profilc
c. Bandwidth or diversion
d. Attenuation
e. Numerical aPnrture
2) LED or Lascr'diride light source
a. Radiation pattern
b. Emissionwavcleng$
''l a,'sPactrallinswidth ' , "' . : ;

a. ResponsivitY
b. Opcrating wavelength

,,

h
\#
i€
 i ,T*io .t.
',t ' f,lc.

Optlcal Communlcdllon (EC - 1 4O2) g7*

c"$peed ,

" d. Sensitivity

Systcm Considerations: .

t : A link power budget is carried out, first we decide the

wavelength and then choose components operating in this '
region.
' 't If the transmission distance is long the attenuation and
dispersion are lower.
* We reset inter pelate the system peiformapce of the 3 major
optical link building blocks such ag, reciever, transmitter and
oiticat fiber;''If'thE c.gmponents ha'ie been over or under-
specified,design iteration needed, First. select the photo
I detector then choose an opficalsource and,see how far data
-
can-be transmitted over a particulhr frber before a repeater is
needed in the line-to boost up'15. power level of the optical

: In selecting a photo detector, th-o rqquirenient is to determine -

the minimum optibal powe? that must fall on the plrolodetector or

.to satisfy the BER. The syCtem parameters involved in-designing
betwee-n the used of lED and Laqel diode'are sigpal dispersion,
data rate transmission distance and cost. In addition to the
attehuation of ,fiber,i6elf, the txcess loss that results from the
cabling process must also be considered ?

15. (a) (ii) Dlscuss ln detall'rise - time budget 9nd liak power
budget psed to evaluate the l[nk perbrmance, (10) .
Ans: Rise Time Budget: :

. Rise time budget analysis is:.a method for determining the

dispersionlimitation of 4*ofiical fiber link. Ttxi system ri:Jfr:
can be determined from rise'time oJ each component $uct-
transmitter, the fiber cable and the receiver in the system.
 tl t'--t-':-.
:-:'lY.t:'
'i-t t
'-"=.$.t

38', B E./8. Tech. Degreo ExtilrlueUon .(Apttl/$t[ay eet O]

For a step input voltage, y tho output voltage from the RC

,.L\
,(
The rise time T. is defined as 10 to gATo rise (or Self) of the
.

T-r = (tn 9)RC = ),.2 RC

The transform function of the circuit is I H(rp) t .

r
.irtt^ iaC Vo ,

-v ------L-
t 1+
R.? iarRC
:- jrc '

=
l+l (1 + *2 Rz CIL/z (1 +'(turf f nz df /2,'
The electrical bandwidthfor the circuit is at which
i

1,
i

H(q) =
l
#.
t:m
P 1

.
--u
.-tFllnvlE,

Total system rise time, TSt.

:Ts, 7'1.1,14( + Atoa + o1,r*
^4)L/2
 ,-=1.s. r.@!B
''S.
Optorl Communlcallon (EC- 1 402) 39s

AtS & AtD - Source and detector l0 to 90o rise nmerespectively.

In the case of Re format, bandwidth and maximum bit rate are
cqual, hence fr B = 0.35.
i
I

Erhax)
ry
In the case of NRZ,
'
2x0.35 0.70
BT(r.nax) s 2B = --.lf =.:f.
'r
Thus the optical communication systems are to be designed
'
a r a t t t .t

such that the rise time should.be below this value,

Ts
r
for RZ
T(max)

0.70 . NRZ.
.
#for
*T(max)
- .l
| \rrrs-fil
r
' t-

Among lhe RZ and NRZ formats, th? NRZ- format. is

commonly us'ed because it permits larger rise time at the same bit
rate.
Link Power Budget:
The irnportance of power budgct is to have enough power at
ttre receiver so as tomaintain reliablo pcr formance during the tife '
time of the entire fiber optic system.

Pr, * P, *in + Ptor, * Pru

where P,loss - Tgtal power loss produced in the fiber optic channgl
' Pst*t
 ,1r*{e{' -t-
f.

"S+

The system margin is the link power margin which is normally

added wilh the total powqr loss in the analysis of power budget.
.t

The link loss budget considers total optical power loss that is
allowed between the source and detector. This loss is due to the
cable attenuation,:onnection loss, splice loss and system margln.

Plor, - aL + oC + oS
v

,,..
a s Channel loss
nector loss

(or)
15. (b). Write ghort notes o:n
: G) WDM. (E)

Ans: l)WnM:
A different wavelengths are cornbined onto the same fiber
simultaneously is known as wavelength division multiplexing or,
WDM. The wavelengths in WDM'muit be properly spaced to
avoid inter channel interference. The main system features of
WDM are
i) Capacity, ii) Transparency, iii) Wavelength routing, iv)
Wavelength switching.
Oprational Principles of IVDM:
The signals from different light sources use separate and
uniquety asiigndd, Th#-wili increar. in the information-iapacity of
a fiber. WDM is same as the FDM.

i) Unidirectional
' ii) gijdirectional.
 .:.'qtffi!r.
-ta"rn-
'.F
- 1... R.
4I''
GptbaI €ommrnlc*tlon (EC-l dt0z)
. . i .,

i) Unidirectional.

Channelsource b,.
Channel 1' I

O\harinel 2 Channel 2

; Fig.:UnidirectionalWDMSystofi '-" :

The unidirdctional WD M system js userd, t9 =cl1nbine, differenl

source. Single carrier wavelengths into.a single frber at.one end
and separati thern into their corre.sponding detectors at the.other
end; A^t the transmitting end multiplexer is needed to Combine the
;;""i,ttg"als from ,"io*
light qources oil to a single fiber"At
tt" r"r.ifrng en{ domulliplexgr.is used to sepallte the s.rgnal inl3'
is
appropriate-bhannet. Crois talk between tw'o:a$acent channels
*gigiUf"et thotr.ansmittedend. The rnain design problem is that
mJri"rtipibxer should provide a low-loss path for each opticaf
Jgn"i td,rce ,to'the ,mgJtiplexer output. Photo detectors are
u":uatty senqitive over a Uioaa range ol wave.'lelgths yhich Tq{
inclirdl all WDM,channels. These to prevent siguificanl."Toyt ol
tte *rong signai from entering eaiih receiviirg channeta Cryi
channel iJOtaf,on o1different wiv.elongths are required, F9r
this
;rrd;,;i*rer Oemuttiplexer must be caretully designed or very
rt"Ut" optical filter withwavelength gutoff must be used'
ii) Bi-directional W DM:
Bi-directionat-finks ire also possible in.WDM systems by
intermixingsou.ces_ and receivers at each- end. This sysJg!
involves*frainginformationinonedirectionatawavelengthl1
 T.,k$... 11.Effi!f.
.: '..!{
'ry.

), s
'; (Aprlt/!&y ml Oi
.f

42 BuEJB.Tech. Dggree Examlnatlon

lnput
$atret

&rSut,
ctranrnl

Fig. : Bid irec'tional WDM eystem.

The insertion loss, channel width and cross talk are the three
t basicperfonnanoe.

' Itis defined as the amorrnt of.power loss that arises in the fiber

It is the wavelength range that is allocated to a particular

optical source. '

It is the amount of signal couplin from one channel to the

. other.
tr). SONET/SDHNe.lrvorh :' - ,.

Basic Principles of SOI\ET/SDII Networks:

The lack of an internhtionatr standard in the telecommuniea-.
tion industry duripg 1980's led to the advent of a new standard,
F.
'first called synchronous optical networft (SONET) and rates
ternied the synchronous digital hierarchy SDH.
- Feah,rres of the Standard:
i) Families of standardized digital interfaces to accommodate
future datatate increases.
 Trih..q-.' y.- ,

'L's*
B.

Optical' Communlcation (EC -i 4@)

ii) Expandabilitytohigher datarate,'

iii) Inclusion of sufficient overhead channels to accommodati'the

iv)' A base rate signals. in the North

of 51'.84 Mb/s to carry all
America hierarchy,up to DS3 (a4'7363 Mb/s)'
v)
tt Synchronous multiplexing and demultiplexing f9r simple
;;;birin! of ihannets fol increasiirg synchrgnization of the
, nettvork
-rh"standardhasseveral,parttheopticllinterfacethat
specifigs the operating wavelingths, po'$'er etc" operation
l:'I
format, the
6;;ifh;r;;s- ind,the-ra6 and format'specifications
ftame size etc.
soNETdefinesasynchronousftamosfuctuqeforsendingthe
multiplexed aigitaitraific and operltiol lI9:1dI* "1": q":1
fibers^. The balic buildinglblock of the SONET has
a bit rate of
51.84 Mb/s.
The corresponding optical signal is referred t: -?n 9C - 1

*t rr"'OC stanOs't for iptiial carrier. The basic building block l is

iriii"ra-r,"* il,i"" ii tss.szMb/s and is referred jo aISTM-I
*t stantls for a synchronous transporl fnod:tler By these
"i.=Str-r
modules,more number oi Uitr can -be sent through mone than
100,00 channels.
I
t
"ls
s.

t//

OPTICAL
CO]VTMUNICATION

S.O. GRIGE $HOBA

. Departmentgf ECE
Velammal Engineering College
Chennai

V"roter?t
November / December -2010 t'i 36
t4ay lJune-.- 2011 I-33
 * xxWw
.J,lt

B,
*.

Seventh Semester
1 ' Electronimar6|..,: Ga1r11n4;f.lFtton Englneering
EC 1 *Oa.OF?rcAL COMMU N ICATIO N

I (Common to B.E. (Part tlme) Slrdh Semester Regulatlon 2005)

TIme:ThreeHoule Maximum: {00 Marks
AnswerALLQuestions
PARTfi* (IO,x | ='20 Marke)
erentiate n& &.s;dM*t$-r+fibers-
: 'r f[I'
Ans:
Mono e'tr'iber Mrrltlmdde Fiber

11) Core radius is small.

'
. .,.,j.

Q) Supports onei m.gde..of Su@ts hundreds of modes.

(3) Losses are minimurn.

(4) For long distance

comm unication .

(5) Supports larger bandwidth.

2. Mention any twai fimfhtions of-Fiher Opfical Communication

Ans: Limitations of fiber optical commuriication system are:

a
. Not suitable for short distance and low bandwidth
:l
applications.
o Fiber splicing is.expensive.

. Susceptibility to physical damage.

 '.rr$
R.

..Ansl......,':-..1.'.l:..
, It is defrncd.r'tt inechanisln that'tates place within tho fiber
resulting in broading" of transmitted light frrtr"r as they travel ,

r- -, '. ll
Tlubtypcs are Intcrmgdal and--.^i.--.^.
Intramodal Dispersion.

, A.ns:
Group vclocity is &fined, as tho vclocity with utrich the

r I dl\-r
l#)
"=
'5. A double.betcroJuncfion InGiAsp LED emftdng ai.e peak :

,Ars;Glvcnr,=30ns '. : ',' l

r rDr:.=lfr)zs.' "r
.'. brrlk recombination time is calctrlated using
rT
'f 'nr
7 :
'- --:-=-
t-*..t
rf- nt

30 x 100
b n
-30+100 l

. . L.

. 6. Comparedircc{andindirecthandgrpmnterial.
 a.

Optlcal ornnqnicarlon (EC-l4o2)

Ans: Indirect Bandgap Mnterial Direct BanlgaP Materigl

1) Direct &ansition is Possible'

from,v4lence to egnduction
band,. ''

2) Electrons and holes have Different mgmenPum values.

same momentum values.
3) ' ExqFpIe: GaAs, InP, InGaA Eiampte: Ge, Si.

7; Why:do we prefer silicon to make lihr opticalreceivers? I

Silicon is preferred as :
o They generate less noise than germanium based photo

,:
8. In a 100 ns pulse ,6 x.106 photons at g wavelength of- fflO nm
faU o, photodetegffi' O" tp-"tolg,5'4 x.106
-, "r'foCaAs
elegtfon-hole paiis are generated. Find thg Quiltum
efficiency.

No. of electron hole pairs generalecl

auantum Efficieoct =

.r*: 5.4 x 106

= 0.9 x lAATo

= 90Vo.

responsivity of aPhotg$iode,
 .T+:a" xt@f
(F
s.

It is the parryneter that defines the photocurrent generated

per.unitopticalpower.Itisrelatedtoquanturnefficiencyas.
I-
ft: ;r#,
o

10. TYhat are solitons?

Group Velocity Dispersi-on pulses to (GfD) causet.*9li

broaden in dme as tirey propagate through 'an optical fiber.
However a pulse shape known as solitons takes into account
the non-linear effects of silica to overcorne the pulse broadening
effect:

PART B; (5 x 16 = 80 Marks)

ll, (a) Consider a fiber wittr at 25rl4m core radius, I core index

-
\ ,(i) If t, = nm, what is the value of V and how many
132i,0
d modes Propogat€ in the fiber? (5)
.ta'l\Jr!-vv Y- vllvl---- -- ---

^t ^ r I3..r-'

(iii) If the core.cladding differenceis reduced to A =

0.003, how.many modes does the frber support and what
fractiorr ofjhe optical
:aI lxlwer f in the tf"aCIng? (O
Power flows

Ans: Given [-0.01

L n1 - 1.48.-

Core radius :25 pm:

I
 x%r.

.#-,.:
t

i) l-1320h.m-given
., :

noa
Y - 4n',,
T fL \t^)(2L'l/2
'
?in x 75' x'19-6 x Q x 0.01)L'2,,,* 1.48
=

:12.453.
,t

f siq Ft+ "

L
,"..:
.o\ Paaa4 4 n tF
ii) -#=:
r =::0.15
3!n 3v77.53
-

j = l|Vo of the totat power lanuched into the fiber.

ffi) [=0.m3ftrven)

rl =*, If i=19;: x 1.48 x Ax o.oo3)r/2

Pc*a 4
' ? --3 tl&.zs = 0,T1,6

-'.- Pcua =' -27,6:% oftotal Power P.

 .--,*p'r.=l%Gj

*
("9

B.E./B.Tecb. Dagrcc Examlnallon (Nov./D*i. zbf0) $

ft
r' '_-i

rtructrrre and'explain ptindplcof working: ,

(l) Stqgtc mode fiber (Step-Index) :

r- l --\

category.

Stnrcture of single modC frber (step index) and multimoder'

fiber (graded index) is as shown in frg. (a) and fig. (b) respectively.

fysarOmsn*rr
--r.I l'fllrh
t$rc&'I
*
d; .* {r} tj 'i*: .-iry tri-f? pfh
fsE)
".*"t-

terry:*ry6r lr1

,:ii'i
rr-'i*
r{ 'r. *..8. .

,/
-..4.'*t.*
,
5-.*.5'

.':
:

i.

r , FiS.: (a) $ln$lernOdeitepindexfiber,

(b) lilultimde gradd index fiber
t,

F
1) It supports only onarnode of propagation,
b
T
&
r

b
 j

1 -'s.
:li

, Opdel Cor-nmunlcatlon {EC-i4OA a

Z) .The refractive.index of,.&e cory is uniform throughoirt and

undergoes air abrupt change at the cladding boundary.
3) Core radius is small.

l. Intermodal Dispersion is notpresent.

2. Mode couPling doesn't exist.
3. Used for long distance transmisston.
4. Higlrer bandwidths. '
:

Limitation:
-I)
Asthe core radius is small in single mode fibers, launching of
opiical power is difficult.
Z) And also LASER sources are to be used which is very
,exlleJlslve.
.t
,

ii) Multimode Graded l$ex Fiber: ,'f

Structure , '\ \

1) It sqpports many hundreds of modes'

2) The cofo refractive index is *1q" 19 ,*ty T a function of
radial distancE from the- center of the fiber'
'i*

1) It is easier to lau4ch optical power into the fiber'

""!r??,#I[ooi,e1"o"..pr"r"i?a.ii#-{'ai""e;i-iii;u-;"
than laser diodes '' r- 'r
,,-
3) Intermodal dispersion is absent due to grad index stntcture.
-,:

4) This allows even higher bandwidths that arc possible in single :

mode fibers.
 .E:4'{'rx%G.
*
(,rs

a'
Limitations:

2) Expensive
3) Can't bq used for long distance tran3mission.
aa

l1.(a) Exptain the.'following with referenee to signal

(i) Bendinglosses (O

Draw appropriate diagrams.

Radiative, losses occur urhenev€r an optical fiber.undergoes a

bend of finite radius of curvafure. ,Fitrers can be 'subject,to two
,b

types of bends:
(a) m4croscopic bends havirrg radii that are large compared with
: the fiber diameter, for example, such as those that occur when

(b) random microscopic bends of the frb?l axis that can arise

These curvature loss effects can be explained by examining

the modal electric freld distributioqs. Any bsund core mode has an
evanescent field tail in the cl:ad in! which decays exponentially as
a function of distance from the core. Since this field tail moves
along wittr the field injhe qqrg1^patt_glthe prop.ag$iq
When "T.try.9f."
a fiber is bent, the field
mode -travels in the fiber cladding.
tail on the far side of the center of curvature must move faster to
keep up with the'fieltr in the core, rs is showrr in Figurg for tlt
lowest-order fiber mode. At a certain critical distan ce xc from the
center of the fiber, the fretd tail woqld have to move fasGr than the
speed of light to keep up with the _core field. Sipc-e this is not
possible the optical energy in the field tail beyon d x, radiates
away.

\
:

F
 Otrlcd Communlcaillon (EC-l 4,tq 9.

/.' .+
/. +

Fig. lsketch olthe tunda rnerffi I mde fleltl ln a curvd opttca I

unye; u lde. (R eprd uced with permission from F.A. J. Marcatiii and S. E.
Miller, BeitSy$. Tech. J :

The amount of optical radiation from a bent fiber depends on

.I

the freld strength at x, and,on the raditrs of curyature R. Since

higher-ord*r'*odrS are bound less tightly tg the frber core than
lower-order modes, the higher-ordcr rnodes will radiato out of the
fiber first. Thus, the total number of modes that can be supported
by a curyed fiber is less than in a straight $ber.' Gloge has derived
the following expression for the effective number of modes N4y
that are gurded by a currred rnultirnode fiber of radius a:

-*F. (+)"'ll
,r

Nqr=N* (1)
I,t
i" c dcfrnes the lradeif-index profilc, A is the core- cladding
rvhere
in&x differoncg, g'is the cla&ing refractive indgx, /c = ?-n/A. is
I
E fu wavepropagationconstant,
r-, and

- q,-:
i
b the total number of modes in a straight fiber.
t
E'
E
E
 ,l.ffir"::lwE
.J,*

B,

IO

12. (eXlD Typcc d'Dtsperdor

IntermodafilqrcrSrin ':'i:r:r, ,'
Whcn an optical pulse is launchcd into a frber, the optical
poryer in the pulsc is,distributed'ov.er all (or most) of thp modes of
, the fiber. Each'of 6o modis that can propagBte in e'riiultimode
' fiber:travcls et,ri sligtitry difffereiirt velticity; 'llis'means:thdt the
modes in a given optical pulse arrive.at the fiber cnd at slightly
,different timeg, thus causjng the prrlse to.spread oui in time as it
traVels along the fibsr: This effdct, *hich is known.qs lntermodal
dspercton or lntermodal distorfion, can be reduced by usinbg a
gradcd-index profile in a frber corc.

Inhmodelllispordon
'-' fi*anoail dkperston ot
chromaAct Asperston is pulse
spreeding that occurs within ri single mode. Tho spreading arises
from the finite spectral emisSion width of aa optical sburce. This
I ptpnomenon is elsq kno*n as group vphetfy &pef$n (G{D),
since the dispersiolr is a result of the group velocity being a.
function of . the w4velen$h, Because intraniodal dispersion
depends on the wavelength, its effect on signal distortion increases- .

with the spectral width of the optical sour@.

,'' ihe tiro main causes of intranodal:dispgrsion is as follows: '

. 1. Matertol D[ryerston, wtrfch arises from _the variation of the

refractirrb index ^of the core material as 'a function of
wavelength. (Marcrial dispersion is sometimqs to as
-referred
chromafic dsperdin, since this is the same effict by which a
. prism sprehd out I spectrum. This causps a:wavelength ,

dependence of the grou. p velocity of any given mode; that is,

'. pulse spregding oectrs, even whin differert, wavclengtho .

2. Vlavegulde dtqperoion, which occurs because a single.mode

frber confines only about 80 percent of the optical power to

t
i

,1i
 .- Fl:-"r-@tE.
&'

Optlcal Communlcatbn (EC-l 402)

the-core. Dispetsion that arises, since'the 2A percenl of the

light propug"tirrg in the cladding travels faster than the light

3. iotermodal llelay: The gther jyjt: Stvin{ d:" t? puly

$preaaing is intermodal de[ny,,which is a reyl! of
each
To1"

, fiequency'. a
 "-rs..t X,
. r,l*
R.

B.E./E.TQch. Degree Elramination (Nov

ili) Given L,=0.5Vo.

. .'.'Atct =0.015ns/Km.

13: (a) Discussthe structure, charactedsfics andworkiugof

(D E:plain Modes and threshold condltion of Laser

diode. (O
Ans: i) LED:
: To be useful in fiber t'ansmission applications an LED must
have a high radiance output, a fast.emission responsg time and

To achierre high radiance and quantum effrciency, Double

Hetero (DH)strueturei are timployed.: . -'

: The most effective of these sffiIctures is the configuration

shown in Figure. This is refe{ed to as a double-heterostructure (or
heterojunction) device because of the two different alloy layers on
egch side of the active region. By means of this sandwich structure
of differently composed alloy layers, both the c?rngr1 axl the
optical freld ateconfrned in the eentral ,4ctive layer. The band-gap
Oim"rences of adjacent layers confine the charge carriers (Fig. b),
t\ while the differences in the indices of refraction of adjoining layeEs
confine the optical fidld'to the central active layer (Fig. c). This
dual confinement leads to both high efficiency and high radiance.

..

C
 n
''rr*
A.

A$t9d.-Cpfprr rnbat ''(Ec-|{oe}

Y.lrl r.mc rr.tlF r-tyr P.w ,;ne I ia6f

(b -,AlrAt (i.Ac
,

--.:tls(l 6rA,t fllr-rAl.Ar (}'l:.Al,Ar | . tutra

rD*l.rc I
r.|!s|Dinar tffilir3x,! l{hlill1 llitd,rrnrr
rdlricr ,Gai.n q646tr iilpr*[Grt
.
crrffica*u co$rcagr
:lrn. - 0-l rra -llrlt
.
-trtr
:- .:: aj .:
:
Lt

trrrict
J-
Gkdrur-htlh
raFrilrrtlnlt
r*.*.+.D l,. . tt$l||lrt

I
-l I
l. I
I
I
,l .. I Atrir. ' I
I r:titn I
I r.- - 'l
I
I
I l:
lr :l
UlW3riie rcJol

Fb. (a) Crosi-sedlon of h typlcal GaAlAs doublq heterostructure llght

emltter. ln thls structufie, r
> y to prodde for both canier conflnement and
optialigutdlng,{b).Eno. ry.bard'dlagrgm,sfrotdng:theactfue regiofi, and '
-
the [email protected], and.,hd9,barrlelg whiah,cpnfine,the charge,carlerq to the
act{elry.er.(c}-Yariatiorrintlerefra{i,veirldexforthe gptlcalgohfinemqSt

Ctarmte.tlhttes :i ', :1 ' :-,, , '

I) terhevelessresponsefirie'('ins);",:'', . :, , ,

2) Optical bandwidth (2 nm or less)

The radiation in the laser diode is' generated within a

Fabry-Perot resonator cavity shown in Figum:.
However, ttris cavity i3 rrruch sma.ller, being approximatgly
25C500,rm long, 5-15 pm wide, and 0.1-0.2 pm thick. These
 Fi:t'':'j'*:'* .-:a,rr,*l!WW.
";'. .,9
s.

r* - B.E./B.Tech. Degree Examintrtlon (Nov.lpoc. zol o)

dimensions are commonly referred to as the longihrdinal, lateral

and trarcverse dimensions of the cavity respectively. ' ) i

fn the laser diode Fabry-Perot resonator, a' pair of flat,

partially reflecting mirrors are directed towards each other,to
enclose the cavity. The mirror facets are constnrcted by making
two' parallel lefts atong naturhl cleavage planes' of the
semiconductor crystal. Th;purpase of these rilirr"rs is to proviOe
strong optical feeaUack in th; longitudinal directiotr, ;h;;t
converting the device into an oscillator with a gain mechanism that
gompensates for optical-losses in the cavity. The laser cavity can
have many resonant frequencies. The device will oscill"te
(thereby ernitting light) at those rmonant frequencies for which
the gain is sufficient to overcome the losses. Th; sides of the cavity
are simply formed by roughening the edges of the deviee to reducl
unwanted emissions in these directions.
\

Cer{y drtce rrc roujh cn

>-l+l*t; -J
Onilcal uutnr*
tri'r*- nnrdhl
intr: l t0*r

.t.

iiinr*s and Threshgld condlfion of Laser Diode:

Ans:
The optical radiation within the resonance cavity of a laser
--
di@,sets. up,a pattern of electric,and magnetic fieldiines called
ther moa"* oitm ca*tyThes"'rrn.conu"niintly u";;d;il i*;
:

23
 "-{idcr-
".ElE%
.,."9
'iI

f COmmurttceti;

in
ii-ii ;;d"s; euct, set of modes can ry srbed
[ffi:tliri"lor4rudigar,ralerl?Tdtranpverse-h,elf-sinysoi{ar
variations of the eleg,trrpqq"qo
i"tdr,.l"$lsth" t-aior axes of the
tt'",d;it1 t :11,:,
eaviry..Ths rongitur$n"1.*our: are iSrliadL of the
-rryQture t"qt'ency
cavjry **.a3,6r*ine the grincip"l e ri i1 mulh larger,
slpc.trrFrr of the ryoiution.'s.iQ.c
"*iri;a'optryur
longitudinal modes can exist' : .
modes
, Lchratmo&g lie in the plane gf the pn iunclig*.,These cavity,
unc ttr3.,wi*th of the
depend rythe *i.aJ1ou ;;*"ration
and defermine _h* -h;#ot1ht
litqil'plqfr': "1 the laser beam'
Trarsverse mode, *
11*o"i*rcd
*ll
itre gtectromagnetic'field
.

and.b*am profite in the directi,on

por,pendi3ular ntye of the
i1lT
'are o1 ,8r"at'ilnpo{tanc:' silce they
pn lunction. These trgdey tli
as
rargeiy determine sugh .]aser chaiacteristiiq ^:_*l"tion
power) and
optical out-put
pattern (the angular disrripurioo.of the

and
To determine the lasing conditions _tl:^resonant
we exprelt tlg.etJcqoryTnetic wlve propagating in
frequencies,
to the mirrots) in
. the longitudinal directiol (brolg the ,.tr normal

is the oPtical radian

 :?&*,61'.'"1wlh.,

B;F-/B.Tech. *
.F

er?aryy .hv' ,g*ryOentifly with the distance z that it

- . -- ^l la

-variel p
tsml,verse s *l on g th e t asin g ctrvitlr accsrdin g to th O re I ati on shi

wnere c i5 the bffective absolption coeffrcient of th9 material in

the opticalpath and f is the optical-Iield confinement factor'

',Las-ingoceurs w[en thg,[qin of'orpor sevelpl 8ui!ef,130es,is

rutR"i"ntio exceed tt e opi"at loss during one roundtrip throwh
inJ cavity; that is, z = Zt-. During this iouhqtrip,'Only ttie frlrgtion
R1 and Ii2 of the optical radiation-are reflecied from the two laser
1 and 2,' reipectively, wher.g and..8a
.8r . m!rror',
.af-9,,,!-!e
"rds
,"n;*i"i,i". n rrr,"r .eneotiort t*m}"nit,'ivt icl iiegiven by
9,

'..^=ffi,].......r,l'
+.

: IQL) = I(0) RrRz exp {Zl{[fg(hv) - "(,hv)ll, ..,(4) ,

At the lasing threshold; a steady-state oscillation takes place,

and the magnituOe and phase of the returned wave must be equal
tothopseof*h:orig1nalwave.Thisgivesttreco$on$......'.
I(LL) = I(0) ,

. , ..t
for the amplitude and
e-iz1t :1
.

c1.,.

t:

l.

t-:

i
 l'l
.,.
ampliftdes can,be found. The condition to just reach the lasing
threstrold is the pdint at which the optical gain is equal to the total
loss asinthe cavity. From equation (5), thil condition is
r .t 1 \
ot:a.;;r,
'tlt

rsth= - * o*d ... (7)

,,
[+] ;=a
wher e a the mirror Ioss in the lasin$ eavity. Thus, for lasing
enidis

. "i'.,: .- ,'

(i) Modulation of a tED (S,

,:
i

The
. r' frequency respo{l$e of an LED is largely determinCd by
.. 1:

the following t[rree factors:

aa
(i) the doprng level,in the active region ,
(ii) thq injected carrier life time r, in the recombination regon,
and

If the drive current is modulated at a frequen ey,o),the optical

output power of the device will varyas :

P(:) = + pr,)21- r/2 . .. (1)

IofI
, . ':' . ;'

ufrere Po is the power emitted at zero modulation frequency. The

parasitic capaeitance can cause a delay of the carrier injection into
&c active junctiox, and, conseqirently, could delay the optical
or{rut. This delay is negligible if smatl constant forward bias is
applicd to the diode. Under this condition, Equation (1) is valid
 18 B.E'lB.Tech. Degree Examination (Nov./Dec. 2I)l o)

and the mo.dulation response is limited only by the carrier

recombinationlme. J

___ vE-. vv !

electrical or opticat terms. The modulation bandwidth r,\aLlr is defined

ro l\IL
as the pgint where the electrical signal power,:designated by p(at),
has dropped to half its constant value ,rrrlting from the
modulated portion of the optical signal. This is the rfJ.ri;;i 3-dB
point; that is, the frequency at which the output electrical power is
lgduced q{
qjB with respect to the input itectrical po*"r, as is

Froqueacy

Fig. : Frequencyrespolse of ap opticalsource slrowing the efedtrical

Since an optidal soulce exhibits a linear relationship befireen

lighq power aod cuffent, currents ratheril;r;;i;;g;; (*ht;h;;;
used in electrical systems) are compared in optical systems. Thus,
tllt: p(ot): PO)/ R, the r3tio of the output elecrrical powe r at
the frequency ot tothe poweiat zgromodulation is
1

.\

l',
t.
 1g
Oprtcal Copmunlcaiion- (EC -1 402)

is the electrical current in the detection circuitry.

The
where I(r)
rt ocuers. at that fo"-guency pornt 'whera 'the
detected electrical powe r p(u) : o+. This happens when
l

1
=1 ;':. (37

:
Someri*"*, the modulltiol-bandldth ol * ltD is grvSl
1"
terms of the 3 dB bandwidth of the modulated optical power P(ar);
p('l = Pvt?'11 tfis
iil*'ir;il- specified at the frequenc{*h.9Tfrom th: ra.tio,of the
;;;;;-il 3-dB bandwiqth is determined value 'of the
optiiaf po*"r- at frequen cy @ to the unrnodulate,4
opltar io*"r. since ih" oitected cufrent iq directly propo$ional

,...Ratioopti""r_10logLffi]=10logH...(4)
:'that
fifie',optical 3-dB point occurs at frequenc"y where the
*
,atio oittt currents is equat to ll2'
" .)
ii) Quantum Laser:
In a standaio DH stru'e (Double Heterojunction
.,L-

,,layer is then (r - ? plq) to confine electrons

iiie
;;fihe "rti+"
fretd, t# electronic and optical properties remain
"il*rr
the samg as in th" bulk material. This limits the achievabtre
thresho1d.'*;n*:aa"*i,.l.*'J"Iationspdedandthelinewidthof
'-
.t-

by having
euantum werl rsers overcome there limitallgnr .an
ness of . arounO t9 n*:. tlit .changes
the
electronic. and optical 'properties dlamalicallv
becluse the
oirr"rri*ality or free eleitron motion is redrrced from three two
 =E*{*,-.'"tl%
,,s,

.:],i .1- ,!i|:-'

tr ',
-:_:-::-:-:;-:.'-'-'"-'''':'1
uqiction' S c-arrier
ttro,d"irnertsior.rs.'As shown in'figure, the',1e.
'!e
;;;, normal to-active large risults in qqanlatisation of-e4,e1gy

f,

L
t-
F
l?
::

F
kr
1,,

t
ll
:,

'l
F
E

\/
\/-,He"[erojunction,:*: ,
t
:: r. ,
,,^^^-^:..-^r:,
ts.
I
:.,,. , _ ,,,

E
The possibte' energy-level , transitions whic-h lead to ,photon
emiision ,t" designated bY AEii'
.: Both Singie Quanhxm
*-"ti (soy) ario Multiple Quantum
'r€glon$,,
structures have single and multiple active
respectivelY.
Active
\
ff,lfuc*on -l L
I

_l arnn;i rrf

Valence '
.a:

band
Fig. : (b) Multiple quanturn uuell IASERS
f"

-."
 .-=,**"" "i.
,

t
g.
'rF

,
In this MQW structuro, the layers separating the active
regions
v are called barrier layers.
fhe MQW lasers t a beti", optjcal-mode connnement,
"n,
which results is a lower threshold current density.
The superior character of MQW devices over Conventional
DH Lasers:
1) Lower threshold currents
2) Narrower line widths

4) Low€r frequency chirp

5) Less tempera[ure d"pq*dence. :

fully explorpd for

Hgrwever, these advantages have not been
longer wavelength MQW laserS using InGaAsP/InP material
sysfem i' -

t4'(a) (if of a photoaetec,$

3ffiffi;Hffi;1"i#5i#Ier
Ans: Pho Modet
to detbctori Receiver Model ;: '-

The sirnple receiver model and its equivalent circuit shown in

figure. The photodiode has a small series resistance Rr, a total
capacitance C6 consisting of junction and packagtng capacitances,
and a bias (or load) resistant R1. The amplifier following the
photodiode has an input capacitance Co and'a resistance Rs. For
practical purposes, R, is much smaller than the load resistance R7

t &rt vrlhage

Photodiotle
.-+-+
hv

. iut rft
-l

Fig. : h) Simplemod"l*_tj:lgdetectorreceiver; and (b) Its

quivalentcircuit
 B. E./B.Tech. Degree braminptlon' (Nov,/D,ec, 2Ol O)

14. (a) (ii) Ilerive photodetector noise of a fiber optical system. (8)

In fiber optic communication

systems, the photodiode is
generalty required to detect very weak optical signals. Deteciion
of 'the, weakest .possible optical signals' requires itrat the
photodetector and its following amplification circuitry be
optimized so that a given signal-to-noise ratio is maintained. The
power signal-to-noisL rafio SfN at the output of an optical receiver
is defined by

signal power photocurrent

N , .photodetector noi$e power + amplifief nolse power

The noise sources in the receiver arise from the photodetector'
a^'
noises resulting from the statistical naturg of the
photon-to-qlectron conversion process and the thermal noises
associated with the amplifier circuitrTr.
To achieve a high signal-to=noise ratio, the foltowing
,ar

1. The photodetector must bave a high qUantum, efficiency to

generate a large signal power.
a

2. The photodetector and amplifier noises should be kept as loyr,

trl.possible.' ' .' ', , : " --'
If a modulated ;ig!;l of optical power P(r) falls on ihe
dctector, the primaryphotocurrentipn(t)generated is . ' .,

'iooa)z
-T ryo ...(1)
The primary current co(lsists of a dc value /p, which |s the
average photocurrent due to the signal power, and a signal
; corriironent io$|. F<ir pin pliotodiodes. the mean-square signal,
.. ., l
current o6> is.

... (2)

6
 .=+*s:*iffiEffir.
.-*
"e,
2I

whore,o is the variance. For avalanche'photodetectors,

I

where Miis the avrrrge of the statisticaltry'va4{ng avalanche gain

I
gilen ''ts u
' : f:,p For, a sinusoilaty varying input dignal of

*o o;iiaei *,the signal component' i.*ris of the form

.$qr1, =
4: *q
.,\
... (4)

no.internal gain are :

(ii) dark-current noise genet 'ated in the butk materiat of the

photodiode, and ' :

la

Therquantum o1 shgt noise ,:r:t from the statistical nature of

the proaocdon and collection of photoglgctro},whyn.an gptical
signit is incident on a photodetector. It has been demoffiated
that these statistics follow a Poissol process. . ., ,,

The quanflrm noise current trry a meail-square value in a

bandwiOttr B which is proportional to the average value of the
photocuxrent Iu: . T

'6

where FtM)is a.noise figure associated with-the rlndof :uture of

experilnental resolts, it has been
lproximation F(M) = M*, where x
(with 0 s x s 1.0) depends on the material. For p
rial. 'For Pin Photodiodes,
M and F(M) are unitY.
 .'B+{ef,'t-
F [FE,
) - r..*
R.

B.EJB.Tech" De3r,€€ Enminatlon (Nir"JD"", 3191

(Z) Dark CurrentNoise

.t , , The p{rotodiode dark currelt is the current thl:Tqnu31
to
incident
now,tUr&ch the bias circuit of the devicewhon;no ligbt:is.
and surface:
on the ptritodioae. This is a combinatiqn -of Uu{r,
u"rt dark cgrrent ips arises frorn electro4 and/or
il;dtme
,iioles whietr"arc thermally'genelated in the- pn iyction of
the
:to ,tn, 'APD,:these liberated celngrs also get
il;ddtJ;. junction'
L*rri"i"a uy tr,u r,igtr etectric $ld nlsenll-rypn
.,,6 ;;; the avalanche gain mechanism'
therefora:miftiplie{.bv

.Porr' ; JO;= ta1frrgYn , ' "'(6) '

(unmytlptied) d*ect1 bulk cqrrent'

whcle is the pnmary -dar'k
6
,/
(3) Surface Leakage CurrentNoise
'I
Tho surface ila$ gurrent is also referre! lo .o u surface
leakageculTentorsimplylheleakagecurrent;-Itisdcpendenton.
. r"*"&,Cufr*,,tfr*iiiJ**; Uias vJftage, end surfdce araal
An
-' ffi#; i;ffi *ii*riI'1"*;." dark e\t4941 tl:1h1,":-q11:-T:.^'-: away
a gUard ring stfl1"nrie rihi,cloi'struns $lrface lbakagq currents
ffiffie loia,.esirtor. Thsmean*quere valtie of the s.p{lce
da*

*i"r"firthesurfa&leakagecurre&'' .,''''t''
(or)

explain how
l1..o)1i), with a f5rpical experimentat arrangencnt

'
nv{f

F
F:--Fss*

"Optlcat Comfiunicctlon (EC - I COg) 25

A typical optical configuration for spot attenuation

measUrements is shourn in Figure. c .'
'i'

Flg.: AIt experimental arangement for making spot (single

wa vele ngt h) atten uat i o n mea s ure m ents u s i ng i nterf ere n ce-ti lters
!
.,, ?nd ernployingthe cut-backtechni,que
The ,interferenee filters ?re .located on a wheol ,to allow
'of
meisurement at a selection different wavelengths. In the
experimental arrangemeqt shown in figuro, the so.urce'spot size is
&fined by a pinhole and thebeam angular width is varied by using
diff6rent diaphragml. Hoqpver, the etectronic equipment .rtitir"i
sith this sefilp is iimitai to that used, for tire' spectral loss
meaburements illustrated. Therefore determination oi the optical
loss per unit length for the fiber at a particular wavelength is
performed in exactly the same manner, using the cut-back
method. Spot attenuation measurements are sometimes utilised
efter frhr cabling iR 'o1der to' obtain iqforlmation on any
;; degfadation in tfre nter attenuation rasulting from the dbHng
 - r"B
g.
B. E./ B.Tech. D€greb. Exrm lr*tlon {}try./Dec. ZOt O} *.
14,.(b)(fiiiExplaln'spohattenuatlon meesuremenf using an
arrSngencnt.
Atrs: Totat Fiber Attenuadon
A commonly
F:d technique for determining the total fiber
attenuation pgr unit tengh is tire crrt-tack o.Ji*Jr*iJ method.
Figure shows a schematic diagram of the typical experimental
sefup for measurement oJ the qpectrdt'toss to sutain'ihe sverall
attenuation spectnrm for the fiber. It consists of a ,white,, light
'source- usually a tungsten halogqn or xenon
arc lamp. me rocusio
. light is mechanically c6o,pped at a lov frequency of a few hundred
heitz. This enables ttrg lock-in amprifier at the receiver to perform
pharylensitive detegtion. Tte chopped lighl is ttren feaiir;;h ;
:l!M1.,,L. ptisg or ditfraction gprling
monoch**,1-!.o1 w.|!ch
arrangcment to seldct the iequifed: wiveleng.th at wtrich'ttre
attenriati.on is to lfu.measurpd. ilence ttre tigtrt"is nit"r"a Llror"
p*S 1o1rs9d on-t-o the fibcr by means of a iicros.op" oUiu"tio"
lens. A bearn splitter may tre incorporited before ttre R-uer to
prtnide light for r*ing optics and a rcference signal used to
compensate for output power flucfirations.

BiedAoP
,Bs4u
.-9467qr:

Fig. : Atypical experimentat arangement forthe measurement of

spectral loss in optical ffbers using the cut-backtechnique .
 Optlcal, GQErrrr{n Fdlot*' (EC -
it is
Wben the measuromelt !s perfo-fmed on'multimode frbers
,"ry d"p"rdent on the opical launch cgditions. TtoreJore unless
trc fr6"t optics **"rg"d to glve ,thg, steady 'statq 'mCId'
distributiorr "r
tn" fibb-r inpq or a d'ry41ny. fitrer is Yqgd'-tbe1 a
"i
*J;ir*a-ffing derrice is attached tb the frUei wittrin the first
;;;: il fr|j'' is also usuallv put thrgush a ilaaaing .noae
,rri*Cr, which may coirsiiiJ of an S.shaped groo-ve cut i1.tle
;fffi and fi1ed witU gtycerine. This divice removes.light
i"**"a ito the mei criiaing throirgh ra$,ationilto the iridex
ffi.;,iili(.;",rt;ii;teh";-';fi "q,t*iiagxlgrv.::':'jII3
stripper gan'also be iricluded at the frber output end torem?n".any
power lvhi"t,ii ;c+iiO *o- the cop inlo$e {aoglnt
the frber
aoo", tne nber length. This tends to be pronounced when
"p,ic1"r
cladding consistsof a lourrefracting index silicone resin.
' fUet opiieaf pqwer at thg'receiving 9n! oltne fib3r is
detected
*ili:;,a'gi.1l-lor avahn&e .prrotodiedo,-,, In-' or&r. t3 ":ht1q
.

;;t6i*"i ieE l"-u," ph6.to.Oetgctqr-"$uqfsoe is yuallv i'**

*"t h"d to the fiber output end face using epory resin or an indgx
matching cell. 'Finaliy, rhe aect{cal output try'. !h."
pirtJJ;r["ior of which is
is fed to a-lick-tn amplifre1, theoutput
recorded.
.The followingrelaliolshlp for the oplical.attenuation per unit
length a6s for the may G o.b.tained from Equation,
Tle{
' 10 ,^- '0, ./r.
adB = T;rqloqofr "'(r)
L1 and L2 are t[e origin{ and cut-back fiber lengths
E

,"sp".tin"ty, Po,,nd rv2 irethe corresponding:".tp:' oel"l

"ri
p";; at a specrfic wavelength from the origpll- and cutback
Fi
fiberlengthr. ii"n* when LT,,1,CI I4are mea''suged in kilometres,
F
5
n
L;
C.

f.r.tt ote it may be wriuen in'the form:

"t
 .':FA:*a**

'r"B
B.
$
2A

'' *bdB =ffilogloq

V,

frcim the
wnere:[ and yicorrespond to outputvoltage reldings
.originalfiuertengthana,tnecut.bact.frberlongthrespectively.
it !'ef""t icat volLges Yi and Vrmay be directly suqstiftted for
the optical powers f5, and F, of EEration' as
they are directly

pro,po$ionatlothesegptit"tp"T*ii . .r:' ,. '..' ,-'.,

The accuricy of ihe resutts obtainqd'ior' aga usili'th" *t 9
istardglyaepo.ndeulonconstantopticalkiqnqhco,nf
i|lnlarirttb
within the t!er'
;#*;#oitt riquilibrium mode distribrition

15.(a)(i) Exptain briefly Noise,pfects oP Didtsl Transmissioh

SYstem Performance' (1O
.:-:'::r
:" ir'-
''

. Arises whori fte*idrttofm;acohereim tasoris ccitlill6d into a

multimodefibel' :
: -:' i
, Thisnoise isaotpr*e+tinsinglemodelin&g"''' l

1) Mechanical disturbances algng Po linl' *"h, "* lbllf:T:

connqctors, splices, micorbends and source or
detector
counlins *" fi;;-i; *JJ *a spafial frrtering o-f lhe optigf
F
L-r
' oo*=t.lnit prduces tempot'al fluctuafions in the s-peclde
F
:
;;i;;* "rttJr"."iving end ttris creating moda-l noise in the
I .-

t recerver.
source can
-:
al..Ihuutlratipiiein the {requenqfii* gBtt-tal
coherent source forms speckle patterns
fqry:
' ' to modal.'de-lays. Atime
;i;;ft#"i"o"" is gre,ater,than interrnodal dispqrsi.g"tl
time (dT).
 :---r$rl 1*n,
."8
g.

Opttcal Gommunlcatlon (EC'-t4ozl a

The modal distortion arising from interferenci between a
of modes will appear aJa sinusoidal ripple of frequency
.

rt efF SOUfCe
Y=Ot---
at
dv'

Modal noise occurs if seurce coherence time is less than

intramodal dispersion time.
'' ' The performancsof a high speed, laser based
multimodefiber
is difficult to predict, since the degree of modal noise depends

,i

-t:
:

1) Use LED's (which are incoherent sources). This totally

avoids modal fioise.

2) Usc{aserivithlargqnr.rmbesgftrongrtudinalmodes,, -li
3) Use a fiber with large N^{,'$ince it supporti large rrtrniUier of
modes and henoe gives a grcarcr numbei otrp""it"..
4) Use a single moiie'fi,-*'i, iince itsupgiiiionly onemodi and

If a cennector bf splicq point couples Some of the optical

pover from fundamental-pode into the first higher ordelr mode
(O" lf 11 mode), thbn a sfrrifr qalrt amount of poier could exist in
tte LP11 mode in a Short sectiorrof fiber between two connectors
or at a rep{r splice. Figure fllusfiate$ this effest. In a single mode
qrctem,':iridril'noliC cirutd occur in shdi'cdnhectorized patgh
co.ds, in laser diode flybeads orwhen twohigh loss splices r."
' "i "
very short distance apart. To avoid this problem, effcctive outoff
wavelength of them should be below the system operating
wavelength. ;
a
 -.'-l*4{,.*t-
[f.,

"''F B
$
30 B.EJB.Tecfi.Dgsreei* !14?tib[ {illov.E.,,

Repair section
LP,,

I
': FlB..,: R€pelrtet$6ns cen pr.oduce model noise in'asil$[e mode ffber
link
.----a

2) Mode Partition Noise:

ilongitudinal .modes of a Laser giod* i.c., Side qef arg not

sufficisnt,suprpressed. Thip is the dominont noiso in SM fibers.
.Intensity flucfirations can ocqr among,thev$iqrsmodeh in a
'
multimode laser even when the total optical output is comtant as.
exhibited in the given figure.

Relative
Arnplitude
kt
E

{'.

*'

F
 :3rtff,'*i!%,
's
,.8

O$fcal Comntunlcatlon' (EC - 1 402)

Because'of the: outpui pattern of a laser diode ig higtly

djrectional, the light frorn these flrchrating modeq gan be coupled .
into a single mode fiber with high efficioncy.
Each of the longitudifal modes that is coupled into the fiber '
has .a different attenuation and time delay,'becagset each is
' associated with a slightly
different wavelength:
Sincc the,power fluctuations arrlong the dominant modps can
be quite largq:signifrCant variatiors,in signal levels can occur at .
' *ti"oir"t-insysieps'uiithhigtrfiberdigpersion' : l
. .i, {fts power.pcnalty
; . in dB cbrxed.by, laser mode'nltitio1 noise
canbgiven as

x - excess noise factor of an APD.

' 1.,:
Q-signattonoisefeetor,. ,t
B,-bitrateinG.Us.- i: ,,' , , ,. ! ',

D - fiber bhtornatic dispersion in ps/nmkrn: ' ' ''

-l f
- k- mode partition noiSe factor.

=
i. to teep power penalty less tharr0.SdB a well designed system
1 .:shouldhavothequantityBLD o1.0.1. l

more pronounQed at higher,bit

Mode partitio4 noi$e"be-cp.mes

eliminated by setting the.bias pqint of LASER,lbove_thresholdt

. . However, raising thb bias power level reduces the available signal

3)Chirpings ': ' - ''

- . A laser which oscillates i4 a single'longihrdinal mode may
experience dynamic line broadening when the injection current is
 - i-.4,.'"_,*r--lWf..
i,- . ..B
s.

B.E./B.Tech,peiiceExarnlnatl0n''$fofi/Ddc j-20f0)
, .a.,

directly modulated. This line broadening is a fequency "Chirp"

associated with modulation induced changes in the carrier density.
LASER chirying leads to qignificant dispersion effects if the
ao

emission .wavelength is displ aced from zero-dispersion wavelength .

Figure illustrates this,

r :4?iilir".*t
F.
'LF
g.

gped A-orylmffir (EC- t,!__Q2) 33

af
"t
I
I
I I
I
b I
I I
I I
I I
I I
I I
Cft
ll
!l
l: . I
It
lt
a
a

$uccessi
Fig.:: Successirne reflested
:
. signals

1) A clbaved silica fiber end face in air typicalty witt reflectrabout

'r U'47+1'0/ ---':'
:

This corresponds to optical return loss of 1 4.4 dB down from 1

the incident signal.

Polishing the,ends of fihr can creata a"thin surface layer with

an_increased
l"lg"Lryj*:. ot aborit 1.6 which incteates the
a

Techniques for reducing opticat feedbacks ,

f) Prepare' fiber end races with a curved surface. This discreases

reflected light'away from the fiber axis, rfrr it does not re-gnter
o'

:.
15. O) (i) Ilescribc, therbasflc' eoncepts of SO]IETISDH Network.
(s)
s
*::. a,
*).+

B.E,/B.T6eh.Deglre'Examln n (l*O[# C."g0*gt. :

Aas: SONET/SDII Nomork.

With the advent of fiber optic transmissio.n lines, the next step
in thb evolution of the digital time-division;multiplexing (TDM)
alled synchronous optical
rietwork (SO*IETlin other par,"ts,gf theworld. . r,i'
i.,-4,

,,i ' {igure shows the.basic structuro.of a $SlEFft, . This is a

twqdithensional.structure cmsistirqg *#.s,eoltrritsi bi g"roms of
bytes, where qne.byte-is eight_blts. Here, in standaiO SO}IET
ter.rninotory,'a cecdo4 co*nicts aAiacent pieces of equipment, a
llne is a io,nger.link that connects two SONET devices, and a path
is a oompletd end-to-end eotnection. The'fundamentat SOUEf
x frame hasal?J.ps duration. Thus, the transmission bittateof tfte
basic SONET signal is S'IS1 = (90 byteslrow) (9 rows/frarne) (8
t
bits/byte)(125 piltrame) = 51.84 Mb/s. This is called an STS-1
signal, wtere STS stands for synchronous transport signal.
For values of N greater than 1, the cglpmns of the frame
become N times wider, with the number of rows remaining at
ninc, as shown in Fig. Thuq, an STS-3 (or STM.I) frame is-270
columns wide with the first nine columns cantaining overliead
information and thc next 261 columns bcing payload data. The
[ine.and section grrertread bytes dffer ssnewhat between SONET
ad'SDH,, so, thati"a translaticiif:' mechenisir'is nceded, to

8r
 -..?ra{fl_. t_-
I
f.,
' r'lt
g.

(EC= 1. 4o2) 35 si\

Stlcot Smmu$le*llEm'
90 x N columns gJ byles

',,. ',"Synehronou$--
:.: j. ,.BalLod'..''
Envelope
(sPE)

\i
N columns
3xN
columns

Fig. : Basic fonna! or an ST,S.N SO,tlET lrame

15. (b) (ii) Exdrir the olnrational princrlpks of wavelength

divisionnuttiplcxing. (8)

fi6 technology of csmbinin! a nuinber of wavelength-'s on to

the' $ame'"fiber is salled,Wavelength Division Mirltiplgxing
(wDM).
1.

Keytr'eaturesof WDM
' 1) Capacityupgrade.

4) Wavelengit iwitchlng.
Flgure .shours ,for. the use'.'of guch componenls in, a tlpical
WDM link containing various types of optical aniplifiers. At the
transmitting ead, there are several. in&pendently modulatea tght
sourggs,*each ernitting signalg at a u4ique wavelength. Here, a
multiplexer. is,needed to.cornbjne theoe ,optica[ outpub ftrto , a
seriaispectru- of.closely tp"""a wavelength signals:and cor$te t
ttem onto 4 single fiber. At the receiving end, a demultiplexer is
 36.

required to $€parat€ th9 optical signals in-tr, appro..priate detection ,

chinnels for signdl processing. At t-he transmitting end, the basic

design challenge is to,have the multiplexer provide a low-loss path
from each optical source:to the muttiplexer output' Since the
. opticai signats thal, are,,.go!qhll!9 u g9tcrally do ,not emit any
signifrcani amourrt of optical power,outside of the designated.
channel spectral ilddth, interchannel crosstalk factors' are
ielatively u'nirnportpnt ai the transmitting end.

.Optieal Fiber

, ln-line
amplifier

Tunaile
' , Receivers
(could inclu{P
Sources
'l
:
B

._{
tr
 1..

:, .,E.G.E.Tdch; Dcgrc? Eiamhatlorp;.AprlUitay

:
mlr
. . : " .:...'(...
ElgclibnlcgrnrlcominBnlca$ofiEqglnerlqg .' '

'i:;'':ri'E6icoe:optcmeolffiiui$e t{'

(Common ioA.E. (Part-Iime) Sixth Semestcr Reguhtion 2005)

Time:TlneeHourcr,,t!aBk!:l
Maximum:l00Marks

PARTA-(|Ox2=20Marks)
1. lYtst are thc advarhges and disadvariages of multimode
fibers? '
Am: Advantages of ltflriltfoode Flbers:'
i) Since the core radius is large, launching of opical power
- intothefiberiseasier. .

. fl) LED sources can be usedwhich is very less expensive. r , '

' 0 tnrcrmodal,Disperyionispreqent. '

tt) Su,pports'tosscrbaneffidtkr' ' :' '', . 't' '-

lli) lJsed for only short distance transmission .
:-
,

' 2. A shp in&x fiDer has I core refraetive index of 1.5 and a
' claddiig refracfive iqdex of 7.#t., getertriue the soHd

' AcceptanceAngle h pno*1= sin-1 <"? - nl)r'.z = 77.366..

3. If fu afenuaffon of an orptical fiberis 0.2 d8/km, d,etrrmine
thc m@ porcr of r 30 km tong frer when the inpw power is
= lfr)pSl.

t
 -.r-w:",iltr.
' ..*
'q.
.*.

{'q9:.:'..-...-
I

:-

(db tkm\ L :'" I-s

.' . Po for 30 km length fiber = 75.3 x 10-5 watts . '

A4s:Gr,q'TrVeIociffDispersion,(G\rE}:...-o.i.i.'.
'
It is putse . sp{eading that arises ,from the. finite spqctral
_
: -..r.

an'oplical'souice.'It is a reiqqlt of group velocity

emission'*iOtt of :_e -

being afunction of wavele*gth,

S. What- are the advantages of double heterostruchrre optical

/: sourees? . 1

Ans: Aryagep'df Do*ffi IIgft roftmefio* Opfieal structure s:

1}H!g[,radianCe..j]:
t

t-
6. , List tH€ a*rantrgssof quantqm *,eII
i
a

k 1) High threshold current density. d

F:
3) Hieh line width of the devicet
I

3. ..comparedto,thet'o!d@ehete$ction$ctureS..
7. Why are semiconductor based photodetcetsrs preftrr,ed to
other typqs of p-hotodete-cto-rs.
dqs:,Semiconductorbe$g-dpfrotodetectorsarepreferred.
 ..Y
'9.

8. Distinguish rqodal noise and mode pattiry

Ansi, ModelNsise': Mode Partition Noise
Use'single mode fiber to It is the dominant noise-
avoid ihis noise. in single mode fibers-
,.
when the light from
I,t arises It is associated withintensitY
a coherent laser is couPled fluctuatigns in the longitudin al
into a multimode fiber, modei of a rasii diode.

9.. Defrne wavelepgth division multiplgxing.

- the same fiber is called wavelength division multiplexing. .

10..DiBtin$Iish.fufodq4entaIandhigherordersoIitons.
Ans:

ietain their shape through -the interaction of balanging pulse

dispersion wjth thenon-linlai propertiel of an frber
:pri:al ., ,
fundamental solitons and those, ,that undergo' periodic shape
changes are calted higher order solitons. .

A cross-sectional view oj the stat wave$de IoO$ the.{ame.1s

' the cross-sectional view of an optical' frber "cut along its axis.
Figure 1 shows the field patterns of several of the lower-order
transverse electric (TE; moOes (which arp solutipns of Ma;rwell':
equaqions for the staU waveguide. The o1{er qf the mode is equal
 '9.
,'g

B. E./B,Tech. Degree Examination (Aprit/May 2O1 1 )

l. GuidsdMo&s
:' fie plots Show that'the elesffiefields
of the guided,modes are
.-not edtnptetelyconftred td tlre central dielec ,tric slab (i;C;,'they do
notgo tozero at'the gqide-cladding interhce), but, in5tead, they'
extend partially into the cladding. The fields vary harmonically in
the guiding relion of refractive index n1 and decay exponential$
oulside of this region. For lgrrr- order modes the fields are tightly
concestrated near the center of the.slab (or the axis of an optical
fiber); wi& little penetr4tion.into the cladding region. On the
other hand, for higher.-order modbs the fields arO distributed more ,

toward the edges of the'guide and penetrate further into the

cladding reglon.
=c,.-
.,]

TEo TEr TE2

t-

Exponeritiat
I:rdding n'x
It lr ,1, dccay

Ha1s1p31;t
varidtion

-Exponcntial
Lrdding ll2
decay
v
i

Fig. 1 : Electric field distributions for several olthe lower-order guided

modes
.,
in a symmetrica l-slab wave-g u id e
Y

In addition to supporting a finite number of guide modes, the

optical fiber waveguide has an infinite continuum - of radiation
*oOrs that Are notlupp"d in the Qore and gqiAed'by the fib.er but
are stills olutions of the same boundary-value problem. the
radiation fietd basically results from the optical po*rr that is
outside the fiber acceptance angle being refracted out of the core,
Because of the hnite iadius of the cladding, some of this radiation
 .re:.-GF
's,
,:

WbeI CorililnEiEatlon{EG*I *S*' 5

get$ trapped in the claddingr therby catrsing cladding'nldes t.o-

appear,.,,I{r.praetice; the cladding,m@s will be suppresed.by a
f*ry coatini which covers the frber or they wilt scatter out of the
fib; after tiavetling a certaiir distance because of roughness on
the,cldding strrfape.
S.LoskyModes

I *r
ov =-r-= l+
-oy-,*A
'
[;rr'"?)
: I '-
., ':i 1.'
. . i

Usurg the recurrence relations fot f,

an{ ,q , we get fiilo seB
of equatilns
'
for Equption (1) ior ttre positivo and. negative signs.
a a .'
Ttc positive sign ytelds
:_
.l
 ,,

' I ir-.' + t;:; - n (2)

@+@=u ::1# ---

'The solution, of thi3 equation gives,B set of modes "caHed the

tv - .fitnl
W
Lt '
ffi=o .. . (3)

Thisresults in ir set of modes called the HE modes.

.t
j =l , * i ;EH *oou, ...(4)
I
t
v,* ! for HE
ru moOeJ
rrr\ri.'vl,
,

di-1(!ta)
W
withiin the weak
ory that wil

srith an EHr,- 1,;i rnotte (i.e., if HE and EII modes of,

corresponding ,";ai orOer m and equal circumferential orderlvf
rom degenerate pairs), then 'any cornbination of an H Eu a L, nt"
:
rnode with an EHur- I , mmode witt likewise constitute a guided

Gloge propobed that such degenerate modes be called linearly

polari zpd(LP):rrodes, and be designated LPirrtrnodes regardless

!
t

t
i
I

r
F
 . ;*lG.
JF
s.

7
O$icat Gommuntca$8n (EG;l {O2}

of their TM, TE, EH, or HE field configuration. the Rormalized

propagation constant b as t Prttion i.s g.rveT various LPlm
"1.V l:t
iaa}si" frgure z.lngeneral, we have the following:
ra?l -^. ^ l-

2) Each LPf* mode comes from TEO*, TMy*, and HEZ*

3) Each LPr* mode (t > z)'is from an HE, * !, ffi and an

l|..(a)(iii) Determiue the fraction of ayerage optical power fesiding

in the cladding of a step index fiber having a core radius
;;; F r-ri" fil{:ry^r?g ar 13oo and has I
numeiicat aperture
rl of 0t.22. (O
"*
Givena=25pm.

2w x 25 x 10-6 -.
','-^""xA.22 I
r/ -ry(NA)=
1300 x 10-' \
:. \
"lP ,\ .\
'2 : 35.26
\
xx
353 modes. ,\ \ \

.'aa
a
Friction of optic-il power residing in the clacding/P6ad is
a

glven as
D
ftt
cu* :
.::

i
-=+Voof To of P'
-.'...PCloi,d,=,7oof:p=-7.0.9
r '353
where P - total optical power into the fiber.
t. \,

F
 '.*rilIIIIEi;{ -
"=+**..'"ilWF,
,.s,

8' B.E/BT. GGlr Dcgrc ffilbn (lpilmy 2ol I )

struc*::r:*::"i:H:'"im
,. ',1 .1 ,_ .i. ,.
"

1' o) u'
il;il :Sffi
the
I
- numerical alrcrture and the number of guided modes.
/l al\
(10)
Ans: \
,

[.-
Dniut Diincnrknt

--L ti, rrt

(yi*o;atl
' -'f' . t-lf 'riu
lcorc!

MCIoud€ tuP,lndci'ffbcr :T
Graded Index f,'lber:
-Defnition:
- Core refractive index,is made to varyrxr a function of
radial distance from the center ofthe fiber.
Nuherical Apeiture and Number of Guided Modes:
The most commonly used construction , for the
'refractive-index 'variation in the core is the poyver law
relationship: :

r a tl/Z
':-
'' : n(r) : n.h - 2^(r) I

,
}""
nL(l 2L)r'2 =
,
ftL(1
\
- A) - nz
l*'"i
: Here, r is the radial'distance from the fiber axis, a is the core
radiaus, n1 is the refractive index at "the core axis , n2 is the
refractive index of the cladding, and the dimensionless parameter
kI a defines the shape of the fndex profile. The index difference A for
the gfaded-inderfiber is given by
r,

h
 '-Af--' Itr-3,-''tlt (2)

:
. ' The loC,al'nume,ricAl€pertrfre is defrned '

t
Ne(r)=]r,,r'l-,?I.t2oNA$)@forrsa.,.(3}
t0 I
Lr

|-
. where axialNA, NA(0) = (4 - t'z)r/, = n.'/x' ..'(4)

? The number of bousd,modes

a

..: in a gfaded-iqdex fiber is

'a
1,o .

11; O) (ii) cakulste the nr&bcr of modgs sup$lrted by'a graded

core radius of 25 /,{m.84d olprsfing
' I

- at
r-:-,.,-820-nm. The fiber has'a refracfive lndex of
1.48 at the
axis anil a clsddiBg
core InIilD
t;UrT indet of 1.46. Assume a
.r z'lr.
parahlic lndex Profrle. (O
Ans:
t-:

:,"
' .

\
' n2= 1.46
 10

a '* 2 {Parabolic profile).

: A=.2,424.

o
No. of modes' M = * aVo
a, + 2 2

(i) inhomogeneities or defects occurring during fiber

, . manqfacturg, Glags is composgd of a randomly Connected
network of molecules. Such a structure naturally contgins
tegons in,which the molecular detpiry is.either'higlrefo,
lower than the average densityin the glass.
(ii) In addition, since glass is made up of several oxides, such as
SiO2, GeO2, and P)O5compositional fluctuations can oceur.
These two effects give rise to refractive-index variations
which @cur within ir,e grass il;rcex
"r* il;i
aie smau.

25
 It
I'Communica$on GC-1aP
t.o!np*d
with the wavelenCth, Thyy'index variations
I oause.

niyf"iet -type scattering of the light'

: "
For single'component glass the scattoring
loss.at a wavglength
, *ririorg iA bttitirll noi*ations can berapproximated bi ;

,+
Here,zistherefiactivqindex,tgisBolEnann'sconstant'81
and thefictive
,fr"iirli"r.J"o*pr".riUitity of ihe material,
,f;
;;;";;;* T1 is' th'e temperature 'at which the ''density
tne grass as.it solidifres. (afer having
il;;fio*rr"'rror"r,inp ig given 4q
;ffiilinto afiber). Alternatively' the relation
.J 'rt-

dscaat=Y; nrt*nrf r .,, (z)

coefficient'
has been derived, where p is the photoelastic .:
- f^
^--^a
-

[ReferQ.No.|2.(aXi),NovemberlDecember201ot.

that cause sislat distortion

12.(a) (ii) Explatn the different factors
(8)
in SioStu mo*e fiber with:n€cessary er(pressions.

of
For single-mode .fibers, *"r"gr1ide disnllsig: .is of
*

importar": ffiJ.,t*" sful6 i*s oc:ulrns ovaldistribution

of the group delay
wavelengths o1 is;bi;tdd from it "'derivafive

owg

Y 'uzLLor'
:-r-
,, *rvul -. . (1)
=-!r ci dv"
 .,8
E

t2 B.E.iB.Tech. Degree Enmlnation (Aprillilay 2Ol l) '

ta

wherg Dnr&) is the waveguide dispersio.ll., , , ,

\ ' To'see the behafiot of the wavegqide'disper"iotr, consider the

expression of the faetoi 'u.a for theJowest-oider *oA" (i.e., the
HEn mode or, equivalently, the LPumode) in the norrnalized

uo: (1 + Elv. ..- (2a)

1 +(4+ffn
=iI ?
W),
ra

Equation (b) yfeJds, for the H Enmodo,

I

fi)'
. 't

b(r)'= t; (1;+
... (2b)
[1 + (4 + t'f
/412
.: ::

Figure (3) shows plots of this exp:essfon for b and; its

derivative -,.Wand * u,
tuncrions,
Was t '-
:'

t-

Fig. 3: The waveguide pararyleter b and lts derivativbs d(W) / dV and

Vd2(W) /dfflotted as a function of theV number, forthe HEr r fitode.
 w -t,--.@T*{,AnL
..
"B
B,
13

12. O) (i) wtth refrsctive ln&x profiles and other relwant

' : shi&agnd-diqpaidonflatt+nedfibers' (10)

- d" (a) Stardsrd SingteMode Fiber- 1300 nn optinised fibers:
Tte most popular singlo-mode' fibers-.- used-.in
. , telocomnunicatiorroitwp.tg are near- stePinde.X frbers, yhich
' are disporsion-optimized,,fo.r operation at 1300 nm. Th3sl
i3gg-nm-optimizia single-mode fibers are of either the matchcd
cladding * tt depresied-cladding design, as shovrn in figures.
"
Mut.n"?-cladding fibers have a uniform refractive ^index
throughout the cl;dding. Tlpical mode-freld diameters are 9'5pm
ard cire-to-cladding index differertces are around 0.37 percent. In
drr$"d-cladding fibers the claddingportion next to the,qore haq
, lorr", index than the outer cladding'reglon' tlode-freld
diameters are around 9pm, and tlpical positivg and negative index
-

' I
diff"r"nrrs are 0,25 ana- 9. tz pgrcent, re$pectivcly'

Material dispereion depends only on the compqsition'of the

material, or"u"gUide dispersion is a functionof the core r1diu1, lhe
I refactive-iqdei diffe.rence, and the shapB sf't.he refractive-index
. profile, Ttrus, tkel*avegpide dispersio. n ^can Tty dll'a$c{ly
with the fiber.derign parimetgrs.,By creatinga frber with a larger
negative waveguide dispersion and asspmi,ng t[re same values for .

rn.t"riul diqperqion can then shift thg. ze1qf-isnersiol roint 1o

._ longer wavclength$.. The iesglting Optical frbers a1e knoqt as
disporsion-shiftedfrberu ': 'r 'i

(c) Dispersion Ftatt+ned Flbers:

An alternative is to reduce frber dispersion by spreading.th.e
.dispersion
minimum out over a widel range. This approach is
frnown as dispersion flattening. Dispersiort--fl{tene0 frberq a1e
' -or" compleito design than disperiion+hiftedfrbers, trecarise the
;ilpers;;**t-u" ionsidgled over, a much,brrydq,,tangc of
wavelength.
 --*qa-r.*reB.
i.g

{tr

14 B.E.lB.Tech. Eq;iree exanllnafidn tnpfnlffi.ryt I)

I ' gtll

I;
JL!.
.|r
.1
.. . :: . i
Fl.
l-
Llrir rr r'l t I - --rr Jr

' . JtIg I s.tl .rr rrl r. trtf

F
 lw%6,
.g

B.

iS

Wavctsl8th (ufiIt
'iai,
 =!w:*r]Q*.
s,
,'i,'S
16 B. E./B.T.ech. Degree Examinalion tAprlUtley Al I )

13t),&nm'opdloiacd

',a.t 'fIJ..
/t-r'
L
b
4
'.!
,. L
l,
g.
\
"iid

sI i,r

g
..!l
u;
ri
*
lJr.
.a
o' ..,r:0,
Dirpa.mioniohlftcd,

lZ.O) (ii) Write a brief no& on polarization tnode dispersion. (6)

StgnaL'which cah v?ry significantly of a fiber- As

{ong
the tttgrl
rt owi in figure, signal energy at a, $ven:wavelength occupies tw9
orthogon"ipolarization poOes. A varying birefring:nce algng ifs
'mode to travef at a slightly
length will r*ur" each polarization
different velocity and the polarization orientation will rotate with
distance. The resulting difference in propagation . times Lr
between the two orthogonal polarization modes will'result in pulse
spreadrng. This is the po;larization-mode disperslon (PMD):

F
 .qr.."{roe.
:
's,

A us-eful means of characterizing PMD for long fiber lengths is

in terms of the mean value of the difi'erential gro,rftelay. This can

(Mprl=Dpruot/T "'(1)

where D *unis measured in PS /rlffi is the average PMD

.:

Fp. T:Yariatbninthe potariatign states of an optical pulse as it.pas-

sest]uotryhafiberwithvaryingbirefringencealongitslength.
 E-.Ery'"fl.qilGf
r$.

Opttel Cqmraunicatio[ (EC-I 4oA

t9

lab,'ix$er34t rypesof SplleingTe${qT,.:,.u . .

FiberSplicing:
A frber splice is a permanent
-typically
or semiperminent joint between
, twg fibers. iiheS" are used to create long optigal links oJ,.
ii'ritUtiors w,[rere frequenlcbnnegtion. aqd disqp,nectign are not'
needed.
!

Splicing Techniques:
', ,Fi,ber splicing teih$ques, j1clud3,
.the ,Fio1. tp*:,,ihe
,

V-gtoor,€ m;chailcal $pliii,. and the-elastic-tube,splice- Tl" first

tefrnique yields a permanent joint,,wherea$ the'otheq two types of

Fusion splices arg made by thermalty bon$ing ,1og:th."t

prepared fiber ends, os pictul"d i1 -fisure 8, In this method, the
fiber ends are frrst pr"uiigned and butted together. This is done
either in a groov"O fiber holder or under a microscope wilh
micromanipritators. The butt joint is then heated with.ln ele.jti:
arc or,a faster pulse sci that the fiber ends a mornentarily method
.and,hence bonded together. This technique can produce
very low
splice losses.

lrbcn
to br spliced

Fig. 8 : Fusion splicing oi optical fibers

 ,r"ffir:.iiQ6:l

S'

B,E./B.Tech. Degree, Exarnination (ApriUMay 2Ol i )

" In the V-groove splice techniquc, the p.repared fiber ends are
first butted together in a V-shaped groove, as shown in figure (9).
They are then bonded together with an adhesive or are held in
place by means of ,a cover plate. The V: shaped ghanne! can be
eithcr a grooved silicon, plastic, ceramic, or metal substrate. The
splicc loss in this method depgnds strongiy on th9 frbcr size
(ourside dimensions and eorediameter variatioqs)' and eccentricity
(the position of the core relative to the center of the fiber).

The elastic,tube splice show.n cross-seetionatrly in figure (10) i9

a unique device that automatically performs lateral, longrtudinal,
'.
and angular alignment..[t spl!.ces mnitimo.pe {bers to glve losses in
the same range as ccimmercial ''fusion qplices, but much less
equipment and skill are.needed. The splice mechanism is basically
a ttrbe made,of an elastic matcrcial. firc central,holo diameter is
slightly smaller than that of the frber to be spliced and is-tapered
on each end for easy fiber insertion. When a fiber is inserted, it
expands the hole diameter so that the elastic'material exerts a
syrimetrical force orthe fiber.,T is.syrnryrefry bature allows an
. accurate and automatic alignment of the axes of the two fibers tg
be joined, Awide.range of fiber diarneterscan,beinsertedinto the
Clastic tu.be. T.hus, the fitrbrs to besplicod'ds not have to be,equal
, in diametcr,'sin@ each fiper,moves into-pqgitio[ indiperrdolrtly
relative to the tube axis.

Y*iooved
subttrate '

Fig. 9 : V-groove optical fiber splicing technique

 r-.,*&*,,.{. "t I Y"mMr.:
(l

B.

OpfrGd@'(EG*l{ } 21,

;'. '

F:
h-
rl

kr*-
!h

&,r,
H:" -
lt*
fl

F
Hi:
s"
F"".
iF':
E__-
r

Si. -

E-
h'
tr

F
E.
r'
E*
r";
ts&

H
E--

ffi
&
 's.

n, s

Fig. 11: Schematicrepresentation of a pin photodiode circuitwith an

: applicd reverse bias.

*ylr*r*tol
el;ctroul
rl:r{Ir.
f.
,l,
^lq .

\aq

Fig .l?;:Simple energy-band diagram fora pin photodiode

As the charge carriers flow through the material, some
electron-hole pairs'will'iecombine and hince disappgar. On the
averago, ' the' Ctrary carriers :mbve a distance Ln or Lp for
electrons and holes, respectively. This distance is known ; the
diffusion length. The time it takes for an electron or hole to
 '*t.F
S.

,O$lcal Gommunlcatbn' (Ee - ilE)

recombine is knOwnr a$ -.ithe carrier lifetirhe and is represented by
:

r,, and lpii,re$pectiv,ely. rne tifetimes and the diftision lengths are

,l

where Dnand-boare thi elgc*o1ryd tro! ai!1sitl

"9"T:"'tt
(g1 constants), ie-speetively, which are expressed in unitq of
centimeters squared Per second.
Optical:radiation ip absorbed in the serniconductor material
according to the exPonential law
-a

. .. (1)

is the incident optical polver level, and P(x) is the o.Ptical power

If the depletion regron'has width w, then, from €quation (1),

a
sorbed in the distance w is
-a llr -

If we take into qc.pount a reflectivity R; at the entrance face of

the ptrotdisae;"tte primary photbcurr*t I, resulting from
thepgwerabsorption of equatibn (3) isgivenby

r-:
Pnv { P;o - i- ',n) 1r - R/ ...(3)

whgre Pr!s the opdcat porrer incidcnt on the photodetector, q is

the electron charge, andhv iq the photon energy.

the difidreni typgs of noiser afiecting the

' 1a.(a)(ii)DesCil,be
Performance of a photodetector and .derive an
efuresdon for the signal in noise ratio. (8) -l
Ans:
[Reter0. No. :l 4.(aXiD, Novembdr/Decenlber 201 0l'
 s,
::{r

(or)
.l

t4. (b) (i)

^4ffi:'
In the high-impedance (HZ) preamplifier design;'theq9?t It
to redice alisouries of noise to the ahsolute'minimgm.,This is
acco.mplished by reducing the input capacitance through the
splection of low-capacitance high- frequency devices, byselecting
a deteetor wi1h.lon, dark currenls, ar.rdby minimizing the'thertnal
. ngisqcontributpd:by the biasin!'resistors, The thermal noise can
b" reOo""O Uy *ir1i a high-impeOaqc-e. aryPlifrer [e.g',.1 bipolar
fiansj+for o. u neia-eftecitnrnsistor (nbf)]rogether with a large
photodetectorbias rosisto-r R5, which iswhy thisdesign,is referred
to as a high-impgdance preamplificr. Since thc high impedance-
produces a large ,lnprrt RC tirhe constant, the ftont-end
band-width is less than the: signa[ bandrnidth. Thus, the input
siglal' is:integrated, and equalizatio4 teclrniques must he

The gircuit of a simple FET amplifrer shown in Figure (13).

le high-impedance reamplifierdesign using a FET

 slsal:Csmmt nlcatton {EG-1402}

twic"lFETs hav.e very large inp-ut resistances R, (usually

' greater-than 10o'Q), so, fgr practical purposeq R, = oq. fiie totdl
i*iim@'R given by gquation ltren reAuces to the va!6'of the

l4;Or(tt) Descrtbettc different error sources afiectingihe optieal

recgivqp;(O - : , ,, ;,
li
.r, :., : '.

Errors in the detiction mechanism can arise from various

noises and dsturbarices. associated with the signal detection
sysryin'fne te1m noise' is; used customariiy tb $gsctibe.unlanfd
rnrnlonene ' of an ' glectric ' signal ' ttat ": tenil ' to' disttfut the
:

ttanimission and processing of t!re, sigr$11apfuysical system;,and

over nrhich ure have irrGcitirplete bontrol. Thetloise sourges'can be
eithai ettsrnei. to,th6,':syst'm : (e.'g., "atmospheric Roise,
equipment-generate$ noise) or internal to the system. Here, we
shaff U" cgncelne{mainly udth intqrnal noise, yhTh.it present in
every coiirmrmication'Systern arrdrepreselts a"baiic limitafion on
the transmiSsion grdetectionbf signals-Thisrroise is caused bythe
spontaneous fluctuations of current Or voltage in electtic circuits.
The turo most comnlon sampbq of theie spontaneous {uqtultions
are shot noise and thermal noise. Shot noise arises in electronic .

devices because o.f the discrete nature of current flow in the

device. Thermal noise arises ftom the random motion of electrons
in a conductor.
Theprrdomaqival.rateof signat:pholonsp:oduces-aqualtum
(or shot) noisq at the"pholodetector. Sjnae this noise depen$ on
the signal Ievei,it-is of pdfticular importance fo1ni1 recgrYerc that
-large
have opticat i"fut ftvefs f* avalanche photodiode
t"t"irs1s-:When using an "ra
'avalgnche photqdiode, an additional
shot'noise. arises from the statidtical natrre.of the, multip-l'icatio*
process. This noise level incieases with ingreasingavalanche gain'
M. Additional photodqtector noises come friim the dark current
an{teakage current. These are independent of '1fo6"photodiode
illuminationandcangenerallybemadeverysmall
t i'-'
;
 -.:?wrE"-G.
's.

"\ s
B B. E./B.Tech. Dqlrob Excminagon Fprt#maf ,Sl, l l

PhototletEctor
Photon streanl ([nin .l{ t
t

i tlull drrt< surrmt Thermd

. Photo detcction o Surfu,:e leakagc
current
, ,q'uantum ' nolsc, , "
I Stetistical gein
.

{luctuation (for
(Poisson
avalrnclte
fluctuatio-n) photodiodcs)

Fig. t 4: Noise sources .rt

e nd, distryba nces in the optical pulse detec-
tlohmechanism

15. (a) (i) Describe the principle of operadon and applications of .

Erbi'm doped.fiber amplifiers with neat diagrams. (10)

- Ans: Erlfu-Doprdtr'lDer Amplifiers

Tre mpst populpr material for lonrg&atf! rclccuhmunicatiorr

applications is a silica fitrer doped,lnith erbiumi ntrich is kn0iun,as
an,erbium-doped fiber amplifier.of EDFA.
'
' foget a urrderst4nding of how an EDFA worfu, wq o&d t"
look at the energy-[evel struiturc of erbium'The oror*um attirnS in
silica are 3* ions, rehfuh are erbium atoms thathavj tct
"rtodE
thrEe of their outer elcctrons. In describing tte transitions of the
oritei'elocEons in ttdsp ionC to higlipr onergr stateo, it,is .iommon
to refer to the process as lta'rsing thc ions to higher energy levels."
Figgry.fl 2)rshows a sigplified energy-lcvel $ag1am and varidtrs
energy-levet tran'sition [rocgsqes of these Ertr ions in silica glass.
The two principal levels for teleconimunication applications arb a
methstable level (the so-called .In4Zlevel) and the. 2pnmp 'the
Il
level, ,The terrn 'imetastAble" means 'thai lifetimes io.
transitions from this ,t"g" to the ground state ar.e very long
comparedwith the li'.times of the states thatled tothis level.
 49A) 27
Op$eal emrltlntcailtoll (EC'= 1

Ground-sate band

is
In norrnal operation, a purnp'ldser emitting-980-nm photons
used to excite i,ons from the $ou1r$ state to the,pump
level, as
shown by transition process f in frgpre (15)..' The$e exci{ed
ions
;;;;t to
fi"ru*j very q"l'yy (in riboui 1ry'l from ll:Pl,1p land
the metastabte uani,.ahovrn,,ffi tralsition process 2. DYdTg this
deCay,'th" exces* itgy is released as plron9.Is,o': eq".'Yalentll,
"t fiber. Wlttrin the metastable band,
ry;;{;ricat vibl{ions lo tt " tend to pqpulate the'lower end of
the electrons of the ;;ti*d ions
ih; band. ,

' Aiiother possibleBrrmp wavelength is 1410 nm. The energy of

-t

thesp pump pt otorr il ntiV ii*iltt io-the T^ryrl-plroton

. .l
elergy,
- r-l^ ^ -:-^l -Lrr.*nrl
Af

photon
b.rt- *lightri t igt er. The absorption -of a 14!0-nT num-p
lightly
excites an-"ir{tron from the gpo,rn{ Btate dilectly. *.o the
;ili;r; 6i-r it ntgstab-le- level, rr. indi.gar:q
"
bY,

proress 3 in figure (15), These electrgns th?: tend to m9y" qoq

"ilifi_::
to the moir*pop.rtaied lower. e.nd ofu the TttTlalfe fevel
(transition 4). SdI oi trr" ions srtUn,g al Pe metaljable llvel can
'absence of an exfry!]lV
ireruy bugk io the ground statE in,ltre
,ti*"lating photorr]flr*, os shown'by transiloT process S..This
il;ffi;iik-*, istnown as spontaneous emlssion and adds to

F
 i !='_lEQ

B.E./B.Tech, @ree Examlnstbn,(Af,W/fiiay 2Ol f ) I

,',,,;ri\ro:rnmc t5@.,of transitions gccur when a flux of..signal,
photons that have energise corresponding to the band-gap ener.gy .
bOtween the ground state and the metastable level passos through
the.device. First,"a small portion of thc external photons will be
absorbe{-by ioBs in'thegrcu$Sptate, wttlch raises these ions to the ,

metastable level, as shown by transitiur'process 6. Second, in the

stimulated emisstid,n process (transition process 7) a signal photon
triggers an excited'ion to drop to the ground state, thereby
emiuing a new photon of the same energy, wave vector, and"
polarizati.onas,tlreincgmjagSignat.pniitoq.''
15. (a) (ii) Explain tt* pOfA architechrre and its configurations.
(o
Ans: EDFA Architechrre:
An optical fiber amptifer consists of a doped fiber, one or
more pump lasers, a passive wavelength Coupler, optical isolators,
and tap couplers, as shown in figure (13). The dichroic
(two-wavelength) coupler handles either 980t1550-nm or
1480/i550-nm wavelength,cornbinatims to couple both the pump
and sign6'l optical powers efficiently into the fiber amplifier. The
tap couplers are wavelength- insensitive wi'th typical splitting
ratios rangrng frorn 99:1 to 95:5. They are generally used on both
sides of the amplifier to compare the incoming signal with the
amplified output. The optical isolators prevent the amplified
signal from reflecting back into the device, where it'could increase
the amplifier noise and decrease its efficiency .
The pump light is usually injected from the samg direction is
the signal flow.,This is known as co-directional Pumping. It is also
possible to inject the pump power in the opp5site direction to the
'
signal flow, which is known as counter directional Frmping. As
shown in figure (13), one can employ either a single pump source
or the dualaump schemos, with the resultant gains typicaity Ueing
+17 dB and +35 dB, respectively. Counterdirectional pumping
allows higher gains, but co-directional pumping gives better nois.e
h
qi
- perforrnance. In additiotrr pumping at 980 nm is preferred, since it
h
p
a'

l'
I
 lt:lqffd-t *t
x.
.f.
i*.

€Pe4 .ffihpmon {BB ;t {ss} .

@s less noisc and.:achi|eyea. larger poHl.ation inversion. than

pttfnpi11ggtl,#S:nm: ": :.'=:i' : : : ::'' :'' .i "';t'"."
a

,.

SiHl
tfi:

sWlrrtr

,..: ; '.'',, , {or1 ,,. r'',,. . ''

rS. &,'(t) Elpm t[i6ONET nqse suucturer and POI-.{ry ilngs
Ans: SOIiIET f,'rame Stt .ucturcs :
'
Figure (17) showsthebasic structurc ot'a SONEI frariro- This
-' . t.
is a two-dimensional struct(rre consisting of 90 columns b; 9 rows
of bytes, whcre onqbyte jp,eighl blts.,,Here, in stand,ar$ SONET
rcrminology, a section conngcts adiacent pieces of equipmont, a
line is a longeitint that corrnects two SONE-T {gy.roes: yg,l!{!
-
is a corpple.te end-tglend col4ectiog. The, fundaqrenlal SONET
taae h:ii a t257s.duration. Thut, the transmission bit rate of the
basicsoNETsignalis l

STS, - (!0
bytes/row) (9 rows/framq) (S bjtslbytey{tzs
1=
pslframe) = 51.84 MUs,

L
 ...,tr
B.

This is called' an STS-I signpl, where STS ,stands for

synchronou{ transBort sigpal, All Othef SONET signals are
integer mpltiples of thisrate, so thatan STS9N signal has a bit rate
equal to N timps 5!.84 Mb/s, When an STS.N signal is used to
modulate an optical source, .the logical STS-N rignat is first
. scram.bled to avoid lon! strinp of ones ani zoros and td allow
1 easiet clock ,rocovely at ,the. receivcr;,.,Aftor,undCi!{ng
:' electrlcal-to-optical conversion, the resbltarrt,physical-layer
opticdlsignalis called OC-N, where OC starids for optical carrier.
In prriCttce,-i1660*coilte cbEfuori to refer Uo SONET links as

j:
. 90 columns of bytes

.4\

'.'".,

.:
becoure N times widerj, withtthd'nurnber'd rows''renlaining at
nt!l& g! $bogE in,figiure{I8},,Thusr,eB STS-3 (or;S-TM.l) frame is
2?0,cellFll8$ wide witlt the first Jrine colurn$s eontaining oveibesd
inform*tion flnd the.gexte6l,columns bcingpalload data,,, ,, :: .. '

'-:.1
 'rr*
s

SBd mr*G$FilN (&*{*oq 31:

90 x NcoJumng of.Pytes

E'
(o
(l)
C.
c,
o
-c
"(,
#
if
\=1 \,i;/
N columns
3xN
columns
87 x N colurnns
. ., .;.. I ' "
: ' ""' " i-'.

loftrmt of an STS'N'SOI*EI frame

, ' ,Fb. l8:Bagio
SONET/SDIIRi[gs:
A kby charfcteristic of SqNEf and SDH is ftat they arc
nsuallyconfrgured as aringarchitecture. _ -^ - -;
The SONETISDH ringp are commonly'callod self-healing
-dngr, since the traffrc flowing along a certain qlth ""I
-auiomatically .

be switched to an. alternate 9r standby path

. following failure or degradation of the.link segment.
Of{re eight possible cunbftRatists o! 11qnry' t& jollowil8 t'w'o
architeanrEshave bellOrr-re popilar for SONE"I and SDH netwols :
o twefitrer, unidirectignal, path-swiched ring (two-fiber
UpS.B):,.' i :;,ir,::. - ;: .-:., !i, ,i: ...i..: :., -..i i=::r* . .^4,'
11,
.. -
. q6rStrr.-o,r,four.Qp*, bidreqlioral; lirle-arvitched r'rng

Figure (1S) shoun a two-fiber unidircctimal.path-switched

rins nitrrori. 6, *rn*tion, in a fnidirectional ring fte normal
*.if,irg tr"m" travets cloclrwise arognt the ring; on the pfimary
pqth. For example,, the cornectiorr from,rql" 1 b-n$d 3 uses
iiritt 1 and 2, whgreas the traffic from.node 3 to nqde 1 traverses
links 3 and 4. T"hus, two communicating nodes ttse a.9pecific
 32',

bandwidth ,capaci ty' areund the _entire,,prirleler of .tho rillg- If

nodes 1 and3 Lxchange inf,orrrration at an OC-3 rate in an OC'12
,tng, then they use ori"-q,rarter o fthe capacity around th3 ring on
all ' the primary tinks, , ,Ir a ,'unidirectionll, ring lh"
pl9t:q1iol
counter-cl@lorise path is used as'an alternate route_ fol
against tintl or noi" failures. This protection path (link. l-g) is
iridicateO Uy dashed linsl.To lcfri9r1pt:t ction,,thg stmalfrom a
transmitting node is duat-teA into both the priTary and protection
'a
frbers.' Thil establishes designated pfotectiol gatn oII, whiqh
traffic flows counterclockwise; namelY, from node 1 to nodq 3 via
links 5 and 6, asshown in figure (20).

Ir
il'

F
L
Y
li

t]
k
;1-
F
H'
F*

frq: Flowof primaryand protectiontraflicfrom node l to node 3

h,
!s"
;dr

E
F:
f
D'-

H
H
IF.
ll.

i
 A^T
r,t
OeSBd Goamunlaation (EC-- 1 {OA

rti Gt€D,Btsqffit,thG qd? of l@M,and' dbnc fo seHeidng''hlg!'

ryocd md high banerHth opdcal scrvi€es. (O

, .,,.' ,freier0.tto.ii.(or(il[hrovelfe/oe9emtq'eorot:'
':
.---.,
. Robof SoHmns:
In an optical corrmuniia{on system, solitons arewrynarros,
high-intensiry optical puises that retain their-shape throlgh tne
irtteraction of b4lancln! pulse-'dlspersign with the non-linear
prqperties of 3n optical',fiber,., if the r-e.la!ile, effects of SPM and
GVD are ponholled jqst right,,a-nd- the ap-oropligte pulse shap is

offso!'the pulse ti@&ning cffect of GVD. Depending on the' '

parriorlerehape choacn, the pulse either does not change its shape
as it propagatos, orit undergoes periodically rgpeating changes in
'

shape. Ihe family of pulses that dongt change in shape aro called
; fundme-ntal. solitqns, and those thlt undergo periodic gha.pe
' c,hangcs are called higher--ordqr solitons. In either c839r
attehuetioo in the fiber will eventually decteaqe the- soliton
energl. Since this weakens the non-linear inter-action needed to
i comte.lact GVE, porit#oally3.spegd qptical amplifrels are. l

required in a soliton link to rcstme the pulse energy.

=-
 QTIESTION PAPER CODE : 5534e
...-...B.E./R;Tech.DEGREEExAMlNATIoN,--
NOVEMBER/DECEMBER 2OI I.
-t.:-r'

SEVENTH SEMESTER
.v.Ltf!r'all

ELECTRONICS ANID COMMUNIC]ATTOhI ENGINEERING

a

NETWORKING I

: r,

Time '':: Threo,hburs ' P[aximum:':,100 merks

Missing.data could be suitably asCumed;
t
-'a

PART A{10 x 2:20 marks)

l, What is'the ener$1 of a single photon bf the light rvhope fu - 1550 nrn,
ineV? ' ' A

2. Assume tlrat there is a glass rod of refractive index 1.5, sur-

a

roundeit try air; Fiird the cliticat incident angle.

3. .Define the attenintion coEfficient of a fiber.=
4. Calculate tlre cut-offwavelength of an optical sighal through a
fibefwith its core refractive iniri of 1.50 unO th-at o'f ,teddini =t .l .46.
The core radius of 25 pm. The normalized frequency is 2.405.

6. Calculate the external differential quantunr efficiency of a laser

,diode operatfngat I.33 pm. The slope ofthe straight lirie'portion of
the curves for the ernitted optical power P versus drive current I is
.a

given by l5 mWirnA.
7. Define 'quantunl efficiency' of a photo detector and u'rite the
.,t
eXpress{Lln. '
.8. Mentiorr the error sourcss in fiber optical rece.iver.
9. What af the three common topologies used For fiber optical
network? Cive the schernatic of any'one netrvork.
10. Calculate the nurnber of independent signals that can be sent on a
single frber in tlle 15,25-1565 nm !and.'lakc,the speciral spacing as,'
per ITU-T recomnrendation G.692
 PART B'-(5 x 16 :
80 Inarks) :

I l'. (a) (i) what'is,,nutr ffure o[ aR optrcal I'r'o(

lt. (a) tii) Calculate ryA of silica fiberwith its'core refracliu: lndex
'..]...,.(d,)of}.48.andcIatldingrefractive,indexoll.46.Wlrat

(4)

(or)
 t

EC24Oz-Nov-Dec 2011)
easurenrenttechniqueuSedinthecaSeof
(i) Numerical aperture
(ii) Refractive index profile
(iii) Fiber cut-off wave largth
(iv) Fiber diarneter. '- ,
(16}
:
,.

15.. (a) Explain the architecture of SONET and discuss non-linear

(or)
-tt'

(8)
(ii) Optical CDMA. (8)
 .F
B.
:

O#tcal Qqmryunicdlgn &.NqUorking (EG3402'lrlqv.Dec 2011) - -

NOVEMBER/DECEMBER 2011.
I
I

TItrI}
SEvENTHSEIv-.IEsT.sR..,-....
ELECTRONICS.'AND COMMUNICATTON .ENGINEERING
EC 1 402,:OPTICAI- COMMUT{ICATION AN t}
-t

(Regulation 2008)
',
Time : Three houns Maxiputn : 100 marks
,
.?,/
.
lVlissing data could be suitahly as$umed.
Answer ALL questions.
PART A-(10 x 2 = 20 marks) ?

1. What is the energy of a single photon of the light whose

eV?

hc
A

6,625x10-34'*3x108 t7
:.. E -=
= 0.0 lZBx l0- J
l55Ox l0-e

0.C128" l0-17
E(eV) - ' = 0.7985 eV.
I .609 x l0- 19
2, Assurqe that there'ir 4 glass roU of refractive index 1.5,
::-air.
suruoqbded by rinJthe critic*l incident angle.
nz
Ans: Critical angle d, = xn'l n1

Where rr, -, rltactive. index medium 1

"^f ::,,er
n, n refractive index of thier medlum

= sin
=4( l-J = 42.o6io ,e

\1.5 I
3. Define the attenuation coefficient of a fiber.
Ans: Attenuation coefficient in terms of dB/krn is detined'as
 :-l*&.idr' 1i@r
.F
S.

B. E./B Tech Degree EXarfrination (EC?402-No_v-_Dec 201 1 )

, a(dB I kml = :;- l0gl

Zs -L ;:i P(z)) i

Where /'(0) -Optical power at the origin (z : 0),

P(- ) - Optical'power at distance 'z' along the fiber:
4. Calculate Ih* cut-off wavel€ngth of an optical signal through
a fiber w'ith its core refractive index of i.50 of
: cl*dding "nJtnat
=" I,.46n The'crore radius,of 25 ;rm. The nor**lized

ZnaNA - Zna
Ans: lcurctlf = :;x
UJ'V..V
-
=li
* rr?,

2 x 3.14 x25x l0{ (t .s)2 - (!40;2

2.405' "'

5. Why silicon is not used to fabricate LED or Laser diode?

Ans: (l) Silicon,is'an indirect bandgap semiconductor,

but not the visible light.

6. Calculate the external differential quantqm efficiency c{a

laser diode op*rutiTg at t;33 pm. The rlNq, of the slraight
line portion'of the curves ior the'emitted oph,qal power P
versus drive curri:nt I is given'by t5 mW/mA. ' ' ---"
Ans:IV*++=$.806i^,,tltm1*d|9twt);1
Y.

= 0.800s r33x14iY!+\
" \mA)
_ | 6.089%
:&
.&:
7. Define 'quantum efficiency' of a photo detector and lvrite
:

Ans: Quantum efficiency M is defrnedas the ratio of flo;'of electron

holepairsgenera]edthetlrenumberofincideritpt,otons
Irlq
ffi:
ffi
K
s
 i
t
..8
B.

Optical Cornrnunica{on & Networking .GCaeq?"Nov-Dec 2

8.' Mention the error souices in fiber optical receiver.
Ans: Error sources in fiber optic receiver:
;L

It can be either external can internal to the systenr. *

External noise - Atmospheric noise,'equipment generated
tlorsg; :

9. What are the three common topoloi;ies used for fiber

optical network? Give the schernatic of'any one' network?
Ans: Three common topologies

10. Calculate the'number of independent signals that can be

spnt on a single fiber in the 1525-f565 nm band. Take the
s.pectral spacing as per, ITU-T recornmendation G.692.
Ans; Mean, frequenct .lpacing as, per ITU-! standard is 0"8 ofil. No.

of independent channels (at 0.8 nm) - .- = 50.

I565 - 1525:40 nm.

PART H5 x 16:80 marlis)
It, (a)(i) y1", is numerical ap"Tury-of an optical; fiber? .

Deduce an expression for the same. (12)

Ans: Numericai aperture is de=fined as the li[ht collection efficiency
a

Numerical AperJur, gr{

Thel meridional ray is shown iri Fig. fbr a Step Index fiber.
t-'-'
; :.
:

F:
h. The light ray enters the fiber core from a rnediuln of qefractive
F,. ,i

index n at an angle 0o *ith respect to fiber axis'and strikes the core-

F..-"
i.l

cladding interface at a normal angle $,

If it s-trikes this interface at such an iingle so that it is totally irrternally

:herr the meridional ray allow azig-zag path along the fiber-
'
core. passrng through the axis of the gtride after:eaclr retlection.

w
 B. E./B : Tech Deg ree Examilatio! ( EC 2402- N or'Dec fl

,presentation of the propagation

Fig. : Meridional ray optics representation prop . mechanism in
-
an ideal'sJep'index optical
l
waveguide
t'

'
internal reflection for the rneridional ray is given by;,, '
*
e 7lz'
sin $rn.,in) = i
:\

r sin e; (max) = tt, sin 0.

n, sin 0. : r,-sin 90"

nt sin Q, = ft=
.L tb

,Y

txz
(D
a

r (: =SlO
t
-l
'|

fl1

n sin 0u,,**, = n_r:]:j?Ou - 0.,) (Since 0. = 90' - 0")

ry
= frr cOS $.

= tll

,fi#
r

l_
I

i
 '&

Qtcal Cornmunication
t t-
I -.
&.Networking
- -t '- '
r -'
(E02402-{rlov-Dec
r - r'
-- -
-'
-,
2011)
I , f
- ' i-
5

nl

NA = n sin 0o (max) lz
r?t -
2
tx2
V

:. Numerical Aperture is defined as,

IVA - lt sin CI, (max) =

,:
given fiber. It is also the figure of merit of a fibpr Thus those iays
having entmnce arigles'0o < On*o, are said tb'be totallyinteryally reflected
atthecore-claddirrginterfacc..,....

11.(a)(ii) Calculate NA of silica fiber with its core refractive intlex

should be the new valqe of 'n,' in order to change the

NA to 0.23.

given fr, - I :48, ffr:

a

1.46

NA=m-g.2,425.
.{i

':l

The new value of n, is given as

(0.23)2 - nr','( 1.46)2 (
.'. n,'- (1.46)?: (0.23)?

n,': (0.23)'+{1.46)2
ff,': I .4780

€
 ...r{
R.

B.E./B.Tech Degree Examination (EC2402-Nov-Dec 2011 )

..:ain.[ephenomenonoftotalinternalrefIectibn
using Snell's law with figureq and calculationq. (12)
Ans: Total internal reflection.
: Refractive index of a medium vglg:tr qL! sl"tlilt_ gL' qf
Velocity of light in medium

High (glass) index n,

ri I
t
I
I
.]I
r'
lncideht i,.-,.
Rav

o ioI
I

I
l"
I
. t' l
L| .

,i
I
'I

Fig,(b): Limiting case of refraction showing the (c): Total internal riflection

of refractive inclices (glass-air) refraction occurs as in f,1gure(a).

.i
-:

It is observed that the'ruy approaching the ilrterface is propagating

 t

*
'l{

Opttcal Co*rrrni.
- at an angle 0, to tlre normal
.l
; at th6 sirrface of the interthce
i. 'z
, , ,lf,dhe dielectric on the. r:ther pide of the interf'ace has a
ret'ractive,
\r
that the ray -
indi,. r. which is less than 4,. then the refraption is-st1ch
path ln it i, lower index medium is at an angle $, to the
normal u4rere $.

is greater'than $,'

' nr sln Qt =
!
fl, Slfl- Q'
l'

i. sind,
-:-:11- r,+

', Some amsun! of tightisreflectedback into the origirytjng

dielectric
metiiurn. (qqtial internal reftection) fi$ure (b),
As n,>n;,
I Z. 0z ,0,.
,itnu when angle of refrac.tion 0, is 90.],.then the refracted ray

, .
This is the timiting case of refraOtion and'the angle of incidence
is

ncmi kl:nun'&fie critical angle S. as in figure (b)'

.'. Snell's law is

. ' ,, , , n, sin-$, ="n, sin 02.' , ' ' .

, $,= 90o, sin 90" = ! . -

' '' sin $" = n,

=:n,
fl2.

'!.?ii' r' .- i;.

: rir.i;.,;:1:1,:":1r;i ]i:l:: ' '1 ,' . I ,, ..;a:, \' -..':: 1"-.
':: : l-, I l
."
t: ,: 0
.
Y.' = Sin-,
-,,
, D'
.1
l\
.
. ,
:
'.

at angles ofincioerice ereafer than .ciitical'angr q.trre lisht is reflqcted

UircL intottreOfginitini;OiateCtiicmertium. ii'-'total iiitg*ai
ieflection

l
 :?is.*t' .t
'.F
B.

cetrying €apacity ,of a fibgr. , .

o
t,
!
e

B
tr
'g
c
t,
0 (c) Bercly di*inguishablc
'4.' pulscs rt time tt> tz
U
.,

-,

(d) Itdi$inguishable
pulfs at timc lo > l!

Drstence eiong I-rDGr

-}
Fig. : Broadening and dtenuation sJ tuo diaqatt ptrlses
as they travel along, a
fibei. (a) Originally, the pulses are separate; (blthe pulses overlap slightly and are
cleady distinguishable; (cl the nrlses overlap significantly and are barely distin-
guishable; wentu.tty,the pulses strongly qverlap and are indistinguishable.
!d)

iF

I
 - ,,.il
!c.
.F
s.
'Opticql Commul,rfFtion
*Iehuorking (E92{0?tNov-Dec201i), .','"": :.,
,,., : A iesultiifdiSpersioninduqpd signal distortion thata light pulse'
is
will broaden as it travels along the fiber. As shown in Fig., this pulse
, 'bip.a&riing will eventually'cause a pul'se to overlap with neighboqrlng
pulses. After a certainamount of overlapfias occurred, adjacent p-ulqes
can no longer be indiriduallydistinguished at the receiverand errors will
occur. Thus the dispersive properties deterrninb the limit of the
information capaciry- of the fiber.

km. For a step index fibr, the various distortion effects tend to limit the
bandwidth.distanceprduceto about 20 MHz kp.i ;.

.' Ih Gradgd lp{ex fibers, the radial reference index profile cah be
.. carefutly selected so that pulse.broadening is mirlimised at a specific
.. . :i,
gperating.wavelen gth.

This has lead to bandwidth-d: istance products as hiigh as 2.5 GHz

km.

-An LED oper*ting at S50 nm has a spectral width of 45

r .../ tt

nm. lVhat is the pulse spreading in ns/hm 4u* to material

dispersion? What is the pulse spreading when a laser
diode having e2 nm spec,trat.*idth is dsd? Ther material
{ispersion is,90 ps/nm krn.
Ans: RMS pulse broadening dug to rnaterial dispersion is 1

o*-.
- mat - o . L .lDru, (I)l

t,

o{rnat) : oi lD*,,(fI
= 45 (nm) x 90 (p'/nm km)

.\

I
 i.

!9., * -, :--
What
, ,. 9:.E.l9I'.[.?==egreg
-EgTil?=!glJq9349?'Tgv;Qge ?0J1)
is meant, by 'fiber splicing'? Explain fusion
!2.(b)(i)
(8)
Ans; A fiber splice is a permanent or semi per:rnanent joint between
,two fibers. Tlre process ofjoiningtwo fibers is called as Splicing. The
splicing used to create long optical links or in situations where frequent
connection and disconnection are not needed. The factors to be
con.siderqdinmakingandevaIuatingsuchspIiceSare:
:

o Fiber misalignments at the joint.

'.o;N[echanibalstrengthofthesplice.
: The splicing techniques that afe generally used:

o Elastic
.'-.
tube splice. t
Fusion Splice
: Fus'ion"splices are forme*by thermally bonding together prepul*d
fiber ends. In this method, the fiber ends first, first prealigned and butted
together. This is done in a grooved fiber holder or under a microscope
with microlnanipulators ;

. Etecrr Arc or
k
ffil
Laser Fusim

r
,t,
*
*rF
!p
sr
f,"

r
X
Optical Fibers
to be Spliced
k
4*
&**
Micro
r.
Manipulate
ry*
Fiber Holders
3
;!ir
,\
r.5
m Fig' : Fusion splicing apparatus of optical fibers
E*
fr-
The buttjoint is tlren heated with an electric arc ofa laserpulse so
F
: that the fiber ends are momentarily melted hnd hence bonded together.
 "&

tj-

Optical Communicltlql 3 Xg!ryothng (EC?402rNey

ieu"'1low.splice'ibsses{TvejcaIIy.ave.raging

:, :
A:possible drawback'of fusion splicing is that the heat necessary
*
to fusei,t. fiber, *uy rveakbn the'tibdf i'n the vicinity of the splice. .'

,,,fr.

t2.(b)(ii) Explafur expantled',beam {iber Gonnector with a neat'

Ans: An expanded-bearn connector, i!lqstrated in Fig.,'effiPloys ,

:,.lenses either.cgllilnate the light

. 'lenses on the endsrof the fibers.
-
T hese
. .. : . : ,

-eryerging
from the tranimitting frber. or focus the expanded beam onto
,f the receiving fib.t. The fiber-to;lens dlstan": i: equal to
the focat tength of the lens. The advarttage ofthis scheme'is that,'iince
the beam is collimated" separation OF tt e fiber ends may take place
rvithin the connector. Thus, the connector is less dependent gn lateral
aligrrments. In addition, optical processing elements, such as'beam b
. -' ,-

splitters and'switches, can easily be inse.rted intg the expanded beam

:, -'-'i:. ..-

between the filter

' ends.
':.

"1: Beam

Receiving Fiber
Transmitting Fiber* -
CoHiryating Focussin g Lenses

Fig. : Schematic representation of an expanded-beam fiber opiic connector

i

F
 r:wru[E
's.

^^,, q
IZ B.E.IB Tech Degree ExaU'latiql (f9ru4
13.(a)(i) ComPare
Ans:

l. Incoherent radiation takes Place

2.
Output Power is less

3.

4.- Wider spe€tral r,vidtlr

,. 's I

-
5. Low cost

6. U-sed at lorver bit rates

'

7,. For shon distance communication

8.
Independent of temPerdture

Life time is more.

*
*

fi.

r
r.,

!,
Ans:
f:

ln this pattern the source is equally bright whe, :i.Y-{ tlotl

; hny diiection but its power diminish:t as cos 0 where
0 is
- the angle. between the viewilg direction and the nonnal to

t
I

I
I

.f
 :.- Fr'1Eec.
,^*
s.

Optical Gommqnication & Networkin( Eg:24A2-Nov-Dec 201f)

Actirc Region :

el

13.(b)(i)Explainthestructure-a+d-rvorki{8o[asiliqon|'PD.

Ans: Constmction and Operation: ,r, , ; :

. [n this diode, the primary photo currpxt is internally multiplied before

it enters the input circuitry of amplifier circuit. This increases the receiver
sensitivity, since the photocurrent is multiplied before encountering the
,

!hermal.noiseaSsociatedwiththete,ceivercircuit.
!

er for carrier multiplication to take place, the photo generated

qarriers must traverse a region where a very hjgh electric field is present.
a

+
n

iP x--
I

,Depletion
Region

;. *.'-
-=!"---J.-----
P,',:

' fig' through avalanche photodiode structure and the electric fietds
in the deqletion and rnultiptieation regions

F
 .-=&!r_."tEeE
-.F
s

. 14 :r; .r,,.r- ..= B

[/B&cl?egr,g.eE*aminqtiqri (:ECe+oz=Nov-oe{e0t1}
=:,regiim
ln this high field
gain enough energy so that it ionizes the bound erectron\ in the valence
band upon colliding r,r'ith them. This carrier murtiprication mechanism is

, ; The newly created carriers are also acceleiated by the rrigh electric.
. field, thus gaining enough energy to cause further impact ionization.
This phenomenon is known as avalanche effect.
when a low reverse'bias vohage is apptied, most of the potential
drop is across the pn. junction. The depletion layer w'idens wiflr increasing.
bias untila certain voltage is reached at which the peakelectric field art
the pn*junction is ibout 5-10% percent below thai needed to cause
breakdown. At this poingthe-depletion layer just "reaches
lvalanche

Normally, RAPD is operated in the fully depletion mode. Light enters

the devices through the p. region and is absorbed in the p region which
acts as the col.lection regioi for the photon generated carriers.'
''
upon be!n! absorbed. photons gives up its energy thereby creating
electron hole pairs in,the p.region, which'are then separat.i uy t igt,
electric field in the p region. The photo generated electrons drift through
the p region in the pn+ junction, rvhere a high erec.tric field exists. tt iri,
this high field rygiol.carrigr rnultiplication takes place.

Ionization Rate:
'
The average numbei of e,lectron, h'ote pairs created by,a carrier '

per unit distance travelled is called ionization rate, Most materials exhibit
different electron ionization rates (cr) and hore ionization rates (B).
^t
There{ore the ratio lr = p/a of the two ionization rates isa
measureof.the.photpdetectorperformance...'
Multiplication Factor r.
t
.' I

The multiplication M for all carriers generated in photo diode is

defined by

In, - averags Value of total multiplied output current after impact

'
{onization.
ar
i

Ip
t.

prim ary uRmultiplied photocurrent.

 l5
ootical Cornrnunicatjon & Networring !F9?a.1tI9.y:P9,:-?9ll)- 'r=
,The,perfor*.n =.===='
resPonsitivitY

8u,,,,,=

'a

where" R;* unitY galrI responsitiv ity"'

,

'
condmilns
il.1u;1ii) Define SA[ ratio of 1.pho,o{*.*:**11:Ihat

;;r, rn auer optic commu,ication'systems, the nlotoaiod: ls senelallv

;;il6 io o"t .t very weat< opticaf signals, feteltion o:.'n: leakest
its fullowt
possibleo Is requires that the photodetector and
r sg, that a,given . nal-to-noiseratio
. srgJ..-. -- - -* :
.l*
amplification ciicuitry be optimized ,"
ratio s',/N at the output of an
is maintained. The pov,er signal-to-noise

.s ,. : Y - :
,ty'. ,, photodetectornoise Power * 4mP -
'lifi"tnoise
Power

nature of the photon-to-etectryn

noises resulting from the statistical
associated wit[i the amplifier
conversrorl process and the thermal noises
a

t)

circuitry.
the follou'ing conditions
To achieve a high sigal.to-noise ratio,

quantum'efficiency to
1. The Photodet' sitor must have a trigh
generate a large signal Po\ter'
should be kept as low
2. The photodetector and amplifier noises

' . r Ft. ' '

gptical frb3r communieatlon

The sensitivity of a photodetector in an
system'is describable :
 :-ria4d._'*-^WS,
.F
B.
-t

B.E./B.Tech ree Examination C?4Az.Nov-Dec 2011)

. ,.'
.
.

Tb see thq inter-relationship of'tlre different types of noises a{lectrng

the signal-to-noise,ratio, let us examirre the simple receiver m'gdel-and
equivalent circuit shown in FiS-
'l'he photodiode has a small series
.:,
, resistance R,., a total capacitance Cl., consisting ofjunction and packaging
a a t.iy tt
*c&pocitances, and a bias (or load) resistor R.. The amplit.ier follorvilrg ^.

the photodiode has an input'iupu.itance C., and a resistanqe R,;,.For

practical purposes, & ir much snlallel tlran the load resistance R, and

Bias Voltage

. Photodrode
4AIAN---------
hv

(a)

Photodiode,

,hv
^//VL-+
t
t
,

Fig. 't {a} Simple model of a photodetector receiver, and (b} lts equivalent

If a modulatd signal of optical power Fft) falls on the detector,

the.prirnaryphotocurrentil,h$)generatedis
"rihdfl::
it,/'t) = -li Pr|
 -

t
tical Networking (EC2102-Nov-Dec
Communication & Ne
al Communimtion
li
2011) l7
'!

mapri*u,:
r-------_i current
-- - consists of a dc val ]re I,,, which is tlre average
. P

photocurrent due to the signal 'power, and a signal component i t,(t). For
p,inphotodiodestlreme4n-Square:,signalcurrent(i.ilis
:\
ti3 )= ol|o,,n=(r;ff))

t:
i

F
I
 -,r=twr..%:
..,*
R.

.: B.E.lB.Tech Exbmin ation ( EC24_2-N ov-Deic 201 1)

The photodiode dark,burrent is thi 'a 'l
'flow through the
bias circuit of the ilevice when no light is incident on
the photodiode. This is a combination of bulk and surface currents. The
butk dark du*ent arises from electrons and/or holes :which
1,, are
thermallv senerated in the pnjunction of the photodiode. In an ApD,
these liberated carriers also get accelerated,ty the trilh etectric field
prcsentat thepnjuneion,'ano are therefore'multiplied by the avalanche
gain mecha,ism. The mead-square value of this current is given by,
,
|.(,3,)=q2ia=2qtii,rru)rr.....]
where /, is the primary (unmultiplied)detector butk dark current
The surface dark cur?ent is also referred to as a surfacg
leakage currmt or simply the leakage curre,nt. tt is dependent on surface
defects, cteanliness, bias raoltage, and surface area. An effective wdy
of reducing surface dark current is through the use of'a girard rini '
i
' structure which.shunts'surface leakage currents away froi itr toui
i
resisto;:- The mean-square value ofthe surface daryr current iqgiven by
I

. l^'
t'et)=ozrt =2q I LB
F \ . ,
I where 1, is the surface leakage current. Note that since 4valanche
h
.t
multiplication is a bulk effect, the surface dark ourrent is not afifected by
the avilanche gain.
p
k Since the dark,gurrents arid ihe signil curent are tlneorrelated, tlre
# total mceil:square photodetgctor noise current (6) can be written as:
P-

.
rji*

h
s
ffi
tsflf
ffi,
(,,i )="i,' =(6li\*).(,8, )
=" b *oLr+ o2as

H
F;,

% = Z,q(In + Ir,)ItfF(MB + uqIP

F To sirnpliry'the analysis of the receivercircuitry, we shall assum.e

H
{'r r
here thbt itte amplihpt input impedafice is'much greater -than tfre f;aO
resistance, so that its thermal noise is much smaller than that of R1 . The
photodetector load resistor contributes a mean-sqlrare thermal (Johnson)
noise iurrent
}
s;

E"
?,-

ry
 's.
*+;
a

. ' -r - lli--l
'
l_ "
"t.
': " '

\'l)='i =1${ a
' I l.
-

,*treie ft is Bottzmann's constant'and 7' is. the, absolute temperattlr€.,

. : :., ;

we get, SNry," at thd iqput of the arnplifier

Whgn, PIn .d@des, aie used: , ,:, - ., " '-

' ,-i' ' I) *tecto'i load resistor (the thermal current i,.}
'' '

..::,.."'2)iheactiteele.merits,ofamplifrreicircuitry(,!,*n)
;

WhenAPD'sare-used:' ,.:. ',.,. '' '

i'

I ) thermal noise is of less importance.

. rl ''

.Z),Quan.tunrnoiiep[qsbutkdarkscufrent.ismultiplied.by.
;'

3) Surfaee leakage current !s not altered by avalanche mechansim.

ma,x'imum
value of M that maximises the SNR. The optimum gain at the
SNR'is givetr ?s,
':

Mti,,
 *
B.E.lB,Tech
[d.("Xi) Explain the filter optic receiver operation using a
I
.gr.tar.=
.-rvE--
(8)
Ans: A sinematic diagrarn of optical receiv'er'is as shown in Fig.
' The three basic stages of receiver are:
I ) Photodetector.

2) Amplifiei
-- -l!-rC^:-
. :

. 3) Equalizer.

t,
i
lk-

Ha,it{0
. :ttflrt*:
P{r1 :

Vout$
'hJt)

',: : : ' Fig. : Sdrematic diagram ol dttyilical'optical receivei

Re - detector bias-resistance, i6(t) - thermat noise due to Ro,.-
& - arnplifier input impedance,i,(t)- thgrrnal noise current due to R ,
FJ
e,(t) * noise voltage source due to therfnat noise of amplifier channet.
F - i!9 :.
ts!!.

.,
j
o Equ alizer tilat'follows the amplifier is a linear frequency
:
shaping filter. It removes the effects of signal distortion and
i nteisymbo I' i nterfere'nce.
'.; .

It accepts the combined frequency iesponse of ,h:

itter, transmission mediqm and {h. receiver and ,;
transforms it into a signal response that is suitable for the ,'
following signal prqqessing ellectrolt ics.

o Let us consid.er that rectangular digital pulses that \,vere sent

.
 C
:\' '&
il'*

Optical Communication & Ilq *

unded and distorted at the
, receiver' the lirrary digltal pulse train incident on the Ploto
l

detector can be described'as: '

' ;
ir'JO'{
/ .,,
I b, hp(t - nT1,l
n=-&

where P(t) '-- received optical power, ::

1-
b..--
,l
An amplitude parameter representing the he message bit,

'" b
*n b can take
-n,---. values of
--'-=.--- b.
-on- and b(Ll
,, conesponding
I -
to a
binary I and 0 resPectivelY,
oa

,T,.
-h - bit period,
' lxp!)' the peceived pufse shape- '
' a

e ' If non-negative photodiode input pulse'is nonnaliredlo have

t-

unit areadescribed as
I ro,Q)dt -l ,then b n

-€
represents the energy contained' in the r{6ulse.

. Tfte mean output current fi'om the plrqladiode { time f,

resulting fro* ln. pulse train Olr, given in equation is

,.@

.A t'' ;'
1O ,i, :

'.

'''?=*Q:'

\q
wlrere Ro = = f€sPonsivitv of'phot6diode.
t.- i;
a This circuit is then arnplified and filtered'to prodtrce a mean
-.\ I

\t

iJ
€
 E xamin ation ( EC2*08-Nov'Dss30 1 1

FET. ', (8)

Ans: High :Irnpedance FET Amplifiers
I

A number of ditTerent FETIs can be used for front end receiver

desigtrls.

The circuit of a sirnple FET amplifier is as shown

':
Fig. : Simple high impedance preamplifier.design using a FET

The principal noise souroes are thennal noise associated with the
FET channel conductanco, thermal noise from the load or feedback
resistor and shot noise arising from gate leakage current. A tburtlr noise
scurce js FET lfnoise.
Since the amplifier input resistance is very large, the input curent
noise spectral density S, k

-++2q Igo,,
S,
"",.i
- t. F.El Rq

= Zq ! eo,, '$
where I .^., * gate leakage current of FET.
gate

ln an FET, the thermal noise of the conducting, channel resistance

.is characterised by the transcolductance g*.
 d{
. .,1{

R.

T
Optical Camrnunieation a Net.utolki!'g tEqz
re noiie peetral density, is

\= t&['
A
8,,

[' FHT chnnnel noise factor which is a constant that accounts ftrr
thermal noise and gate induced nerise plqs the correlation betlveen these
two noises.

The thermal noise two charaeteristics W atthe bqualiser output is

then

r . *#t $
ry, ffi],, {1t)'
I
YF = #[r- sottu-
'

where f=Cr+Cd*Cg_+Cs,

The I lf =noise eomer ftequency{ is defined as the fiequency at

rrrhich lffnoise, which dominates the FET noise at lorv frequencies has
a llfpower spectrufil, becomes equal to the high frequency channel
noise described by I. ".
The high impedenqe arnplifier design technique y'ields low noise,

14. (b) Explain the me&surement technique tlsed;in the c&$e of,
-l

'(it) Refractive lndex Pro$le

(iii) Fiber cut-off wave length
(iv) Fiber diameter. (16)
Ans: (i) Numerical Aperture
. The numerical aperture is an lrnportant optical fiber perameter, it
t

affeets eharacteristies such as the light gathering effic'iency and

normalized frequeney ( n of the fiber.
r "The numerical aperture (NA)'is definbd for a step,index fiber in air
is given by,

NA ='siR 0* = ( ul *
is the acaeptanee angle.
* core?efrdCtive index, n3 * cladding refrae,tive index,
 v-Dqc 801J
'.,!{

B, E.lB.Tech Degrets Examinatlon g.

ie t{-e(n) for a' graded index fib.er

at different radial distance from the core axis may be

"l €ommonly used technique fbrthe deternrination of the fiber

numerieal operate are:
;

i) far fietd angle from fiber using a seapning photode'

F,
tectar and a rotating stage.
,.,
ii) far field paftern by trigopometriq'fiben

Far field pattern of NA meesurement using I rotatlng stage:

tlg. I Flbrr numellcal aperturr mee$uramanlualng t lcannlng photodetaetor

* i r andarotatlng*tage ' ' s

e The fiber output end is then positioned on the rotating stage

with its end face parallel to the plane of the phertodeteetor
inpdt, its piltput is perpendicular to the axis of rutation.

all possible angles using optieal system similar to spot

e The pheitodeteetor me-y tre eithsr a small area device or an

apertured large area dt
fiber end pnsitioned in oder to obtain a maximum signal with
' no reitation (0o).
 r?{F-...i"Etr*r:

Opticql Comeunication & Networking (EC2403-Ngv-Dec 2011) ?5

'&.

r The'iotating stag is turuetl tire limited of far field pattern

.

., o The output power and maximum acceptance angle being

obtained when the power drops to 5% of the maximum
: intensity. '

I The far fielcl pattern to obtain NA usiirg trigonometric fiber

.' ,The ertd prepared fiber is located on an optical base plate or

t Lighl is launched into the fi.ber under test over th€ full range
of its NA, and fur freld pattern from the fiber i"s displayed on
a screen rvhich is positioned a known distance'D fronr the
,\
fiber gutput end face.

I Screen
Lens
D:'--+l

Lightfrom-
SarceOpths .
'j
I
l_

Vibration-lsolated
SlabGranite

Fig. : Numerical apefture'measurement using trigonometric method

,A
^ NA = gin 0u ='
lu, tzl, *r']' t2 (Az + lDz )r/2
where.A is the screen size, ? ir distance of the fiber.
Using these rnethodso wb obtain the,average value of
F.
I

.-

r
"/ r The refractjvl index profile:oh the f,Uei .or. ptays an
important rore in characterizihg the properties ofopticar fiber,. .
and todetermine the nu,n".i.Ilrp"nrr. ,rrU",
";JunJrfr.
of,modes propagat[ng within the fiber core.

The impulse response and the information-carrying capacitS,

' ofthe fiber is shonglydependent on thq refractive index profire.

Different techniques used for the measunement of fiber refractive

lndex protile
,

ii) Near Field Scanning Method

iii) Refracted Near Field Method (RNF).

lnterferometric Method
Interference microscopes (eg., Mach-Zaender, michelson
)

profiles of optical fibrers..

The"techniqug used dn preparation of a thin slice of fi'ber

(slab method) which has both ends accurately polished
to
obtdin square (to the fiber axes) and opticallyr flat rurf*;.
J

, Two major methods are:

a) a trartmitted light interferometer

i b) a reflected light interferometer.

I

In these two rnethods light from the miscroscope travels

and
normal to the prepared fiber slice face (parallel to the fiber
axis). and difference.s is refractive index result in different
optical path lengths. e
j

When the phase of'the incident:light is cornpared with the

phaSe of the emerging liglrt, a field is paraUef interfbrence
' . ,
fr rr
fri,nges is observed.
 timl Comnrunication & ECI{0l.Nov'DBc 201
i mnS !e taken'
t

The fringp displacements for the peiirrrs w]1\in tlre

fiber core
?
and are **u*,.*o using as referenee parallel
fringes outside

thc fiber ebre,:.*

a) a transrnitted light interferometer r
\
' b) a reflected light interferornetel

_)

Flg. : The principle of Mach{ehnder intarferometer

I! ' The refpctiveJfldex difference.between a point in the fiber

(\
lore anO the cladding can be obtained front the fringe
shift
r;, whieh corresponds to a number of fringe displacements.
r

t :The
:-

difference in refractive index 0u is given try :

'.:
aqL
o, = -:-
.r ' ,. ' '

where i .is the thickness of the tjtrer slab. I is the incident optical

1 .z''
. 1 r :

.eii'
 ?8 Erylilqtiql (8C240tr.N0v'Dec
B'E.lB,Jec[ E$req Examination 201 t
ffi heari ng Interferomstry
I

The liglrt beam is incident to the fiber perp€ndicularto its

uxis,trhir is known as trensverse shearing interferornetly.-

.lJ\

(iii) Fiber cuioff wavelength

t A Multimode fiber has msny cutoff wavelenqfs becluse
the number of bound propagating modes is usually large.

For example, consider e perabolic refractive index graded

frber, the number of guided modes 4 it

Mr=[f)'(,,1 -ni)
'where
a is the core radius, tt, iS cott peak index,},, is cladding

1!

The wavelength is irtcreased- a growirrg number of modes

are cutoff wavelength of a
LP, * mode is'the maximurn
wavelength for guided fiber.

When the fiber diameter is E or 9 pm. LPrr mode is the

cutoff wavelerrgth of single-mode fiber operation.
,

When largv ittenuation of LP,, mode effective cutoff :

wavelength parameter is used to rneasure the cutoff

wavelength ( 100 to 300 nm).

The effeetive cutoffl wavelehgtr is measured by increasing

rhe signal wavelength in a fixed length qf fiber until the LP,,

t.-
Transtnitted power techniques recommended as reference
xa
':
test method and anotlrer method conespond$ to EIA standprd
^
I

h, test m€thod.
t. L

r
 ';
Opticdl Communication & Networking (EC2402-Nov-Dec 2011) 29
'T'he
,
r effective cutoff wavelength is defined as wavelength
a
greatir than the ratio between the total porver, to ttre launched
frigher order rnodes and fundarnental mode power (< 0: I
rrlBrrvr VI uv. lf lvvve ts..v
:-"

....---r1,
t''. I

, (a) (b) :

Fig. : Configuration for thlfieasurement'of uncqbl$ fitrcr crutoff wavbtength

s (a} Single turn (b) SPin mandrell

both fundamental and LP,, modes.

I

l.

;power 6nd the fundamental pewer

.a

P' (r)

' P,(f,) is total power, Po(I) is the fundamental mode power.

.. 'The effective cutoff^^wavelength l.* may be
vavelength l"* be determined as
the longest wavelength at'which the bend attenuation or level
I

difference ao}i * 0. I dB shown below:

L.

it:

F*
E
ktk
}.J
f,E!-
il'
ffi
k-
Is.
M,
5ii-.,:
*F*
i&x
*.{;* *i.{r-

:,"4
r&rdli,'

'f''' '"rys.
 B.E./B:Iech ree Examina$orr EC2403-Nov-Dec 2011)

Bend
Attenuatbn
a6(dB)

0.1

Fig. I Bend attenuation againat wavslength in the'bending method for the

,, t
measuroment of cut off wavelength I,.
. The relative attenuation a,,,,(L) or level difference between
the powers launched into ifre rrTultimode and singte mode
fibers Hlay bg,(using power step techniqu€)

where a,(I) is rElative attenuation or level difference,

F-,{},.) is transmitted powerz
P,(I) is spectral power of mullirnode fiber.

Spot
Sizeorg

).ce Wavelength

Fig. : VYavelengfih-.dependence of the spot size in the spot size technique for
the measurement of cut.off wavelength i.,.

i
'

l'he another techrrique to determine cut-off wavelength is

spot size technique. When the fiber is operating in the single
' ; -\
\
 .t

l\
'&
C

.rr..) : ,- .- r,:- '

oryr'ilHm"Tu1#HHlil8i'58',##*H3iirHEilr,--imrru**'il'

.
' wavelength,

'fiberwillbesmallerthantheuncabledfiberbccauseofbend
' *Pt615:(miqo bend macro bend)'
(lv)FiberDtimeter ! .''
the fib€r
r Dudng thefrber manufacturing process, i1 illseltialthat '
''oute.-diamlterjs rnainuined cons[ant to within 1016. Any diameter may '
causb excesdive ,.uai*i", fosses and make accurate
fiber'fi'bcr '
.
o0nnettio0difficulrHenceon.linediametersmeasuQmenla&required
gr€ater
which provido accuraay better thon 0.3% at a -mcasurement

The mopt'corfltroilohltine measurement techniq$ usl

' fiber image
at.1
projection(shadow me$oa). tn this'meth$' a laser beam i$ swep't
,onr,rnt uuto.ig tt"nsverselv acroqsrhe frber.ard a melu3ment
is :
t
the
made ofthe time intirvat dirring which the fiber interc3nts leam,l1d
wavelength is
casts a shadow on photodetector. Thc beam's operating
0.6328 mm and is collimated uses h'Jvo lenses
(Gr and.G:)'
.
galvanometer
Two ririnors M,:and'M, are usod; Mo is driven by
a small angle at a con$ant angular
yelocity'
which makcs it'r"otaEe'drrough
Collirnetor

Gi G.e --+E4dr-.-F_rE

Photodetrctor

':
Flg.i Th.s Ehadow mgthod for on.llnl measurement of fiber meter dlanteter

I
 s
.

32 B:E.rts,Tech Degree Exemlnaton (EC2402'Nov"Dm 1t11)

(C,,) is swept across the fiber by the oscillating mirorarid is incident on

.. The velocity d,r'/dr oftlre fitrer shadow located at the photodetector

:isdirectlyproportionaltothemirrorvelocityd0ldt,.....'

ds l,dl
dt= 'dt
'-!-

, .l is the distenae botvveen tho inimor and thc photodetector.

- The shadow is registercd by the photodetector and an slectrical .

pulse of width (or") which is relatcd to. The fiber outor diameter do

dr
T.
.4i = 0.-dt
may be quickly determined and

. ; :.

Adventages:
l) Morc accuracy. 2) Faster diameter measunements.
15.(a) Explain the archltecture of SONET'end dlscuss nonr
linear effects on Network pe'rformilnce. (16)
.

Ans: SONET and SDH ane configured in a ring architecture. [t is a

F key.characteristic,
::

These SONET/SDH rings are called self healing rings since the
traffic flowing along a qgrtain path can automaticirlly be switched to
an alternate {or) itand bi path follow'ing failut€ or degradation of the
link segment.
Of the eight possible combinations of ring UpGq the cornmon
two architectures are:
{) Two fiber, unidirectional. path switched ring {two fiber
UPSR.). i

2) Two filer or four fiber, bi-directional. line switched ring (t*o

fiber or four fiber BI-SR).
Figure slrows the first t-vpe (two fiben UPSR).
IF*

Optical Communication & Networking (8C2402-Nov-Dec 2011 )

, ,Fig.' : Two fiber unidirectional netvdork with a cot nter.rotating irrotection

path.
, .' By .corlvention, in a unidirectional ring the
lormal traffic travels
clockwise round the rirrg, on the primary path. For example, the bonnection
from node t to node 3 uses links I and 2, whereas tfrl traffic node 3 to
noelel traverses the links 3 and 4. Thus two communicating nodes use a
und the entire perirneter of the ring.

r ----- is used as
rth at1
alternate route for protection against link or node failures.
This protection path (linkss, -8) is indicated by dashed lines.
:Io achieve prot6ction, the signal 'lrom a trarsmittirrg
nodi ir dual
fed into both ihe primary and protection fibers. ",-
.i
on which traffic flows
counter clockrvise. namely, from node I to node 3 via links 5 and 6 as in
(Fig.). TX

PrimaryJ

l! lode

'I
lod, q
f, lode
2
n4
|Ir 'r
1. r
t
T
I
I
la
rt)
i Prtteaion patl
J 2

i.I
RX

Flow of primary and protection traffic from node 1 to node 3

 .:rag ql%.
;,r+
g,
ts.E./B.Tech
?tim (EC2402-Nov-Dec 201i) _*
Consequerttly two identical signals from
their destination from opposite directions, usualry with diftbrent delays
- as denoted in Fig. The receiver normalty,ele.t, the signal rro* tt.
primary path. However it continuously compares the fidelity of eacir
signal and choosesthe alternqte slgnal in case qfsdvere degradatibn or -
loss ofthe prirhary signal.

Thus each path is individually switched based on the qualitv of the

. received signal.
,

F-igure illustrates the architecture of a fourr fiber bi-directional

switched ring.
/
Here two primaly fiber loops (with fiber labelied segnents I p through
Sp) are used for nermal bi-direotional communication, and the other two
secondary fiber loops are standby links for protection purposes (with

Non-linear effects on network performance.

I ) Transmission of different wavelength channels at the highest posiible
bitrate. .1

2) TranSrnission over lonsest possible distance with the smatlest nurBber

ofoptical amplifiers.
3)Network architectures that allow simple
-
and efficient netrvork
operation. control and management.

To meel these challenges, careful design --' be followed, since

3-- to
there are effects (listed belor.v) that affect tlre optical fiber rransmis-
T
siorr links. '

These efTects are inherent in fibers.

\

l) GVD-Group Velocity f)ispersion that limits the bit rate by temporally

zlNon-uniforrn gain across the desired wavelength ratlge of nnFA,s ip

WDM links.
3) PMD - polarisation ntode rJispersion that arises dure to tfue dilferenr
velocities of t'he t\&'o ortlrogonal polarisation rnocles tlue to f.rber
 .

''&
i

& 2011)... . ,, --35

*!g+ CgftTgnlc?t!.q.q lLet!{grki!g-(E-C2402't\hv'Dqq ,. ,,

;:

cause irrstabilitieq in
4) Refleetions frorn sPlices and conlleators that .:
t

5) Non.linear inelastic scattering processes which are interactions.,.

"'
i;;;n;;i*;i ;,*;ali and moLcuhr G0 accoustic vibratiorrs in a

6)Non.linearVariationsinrefractiveindexinasilicafiberthatoccur
-t
changes in the
b..uu*.,ircfraqtive index is depgndent on intensity
signal, -.
Two Non-Lineer Effects
-I) Non-linearinela$tiescatteringprocesss''

nce on intensitY c hanges;

;;ilprras9 modulation XPM (or) cPM

'. 3) Four wave mixing FWM'

signal-intensity'
,"ruurungthehannel that qrr dependent on the optical
Thesenon.linearpnocQssesprovidegainstosomechannelswhile
cross talk between the
aepteting pcnver from others, thereby pmducing

SBsdegradest}iqcarriertonoiseratioifthescatteredpoweris
equaltosignal inPut Power:' '
nottrspuandSPMaffectonlyphaseofsignalswhichcause
.:chirping" in .digital Pulses.
,I
'
This can worseft pulse broadsning,due to depression".particularly
in ve$ higtr rete systems (> l0 Cbls)'

Power Penaltv
slglral loss. all
Wherl any of these nCIl'l-linear elftcts aontribute
tc)

*
 t
r
'1{

I * g*of,
Lufi,*

*
l-e*uL LA
(or)
sL
L,i *total ampl ified link length,
,* lenSth between the amplifieros. .
lpan
'The
, effect of non*linearlty inmeases with intens:ity of input signal
pqyen this intensity isinversely proportional to the eross sectional area
of the fiber sore. Since the power is not distributed uniformly amoss the
fiber eore cross*sectioryeffective cross-sectional area Ar,$ean be used.

(i) Wnvelength rnuted netwol*s. , {8)

iiil Oprical'EoMA,. (8)

o Two problems arise in broadcast-and-select networks.

:* morc wavelangths are needed as the number ofl nodes in the
 l

gglictLHlpunicbtion &Ietygting (!iS,#p2jyoy:Pg3g3]1. . . 3i

large number of users spread over a u'ide area cannot be

readily connected with a broadcast and select network.
I Tlris is beeause of the splitting losses of tlre passive star

a Wavelength - rotted networks overcome these limitations

through wayelerigth re-use, wavelergtlr conversion, and
optical switching. . '

The physical topology of a wavelength-routed network

consists of optical wavelength routers inter-connected by pairs
ofpoint-to-point fiber links in an arbitrary mesh configuration
as illustrated irr Fig..

o V{avglength
routing node

tl End station
. .t-----

Fig. : Wavelength re-use on a mesh network

Each iinn can carry a certain nurnber of wavelengths wtrich
can be directed irrdependently to different output patlts of a
node. e

r L,ac-lr node rnay have logical conrrections will severirl othernocles in

' the nenvork, wherc,.rych connection uses a particular wavelength.
. ProVides the paths taken by any tw'o connections do not
L.. .
overlap, they can use"the same wavelength. thereby the
m.

il number of wavelengths is greatly

:-
reduced.
r-
Optical Cross-Connects
?

h
4
l:l
r
Li'
t"-
E-;
. The concept of an optical-croSS-connect architecture iS the
Ior
&-
E a

F,
achieved.
F

H
$,-
at A u$er site can be
l
F:
r_
ts;

F.,'
F'.

F
 _:=q;ril"r*qilE.

g.
t -'t+

These Optieat Cross-Connects (OXCs) operate in the optical

domain and ean route, very high, cepaaity,'WtlM data stRrams
a

over a netrvork eil'intercormected optical paths,

Tunable
Filter$
$
EDrm w
'-> I

t
InPut
c
fibtr:1 h

m
a
M
t
U
r
x
i
I
x

Dropped
0ptieal
$ignah

Considerthe OXC architecture shown in'Fig. that uses spaee

swite hing wiihout wavelength eonvel'sioR.
.\
The speae switches ean be constructed lof a cascade sf
electronieal ly contro| led optical directional-coupler elements
or semieonduetor:optical-ainpl ifier sw itch ing gates.

t EaCh of the input fibers carries [vI wavelengths,

At the input, thg arriving aggregate of signal -wavelengths is
?-

amplified ard passively divided into N streams by a power

splitter.
.

Tunable filters then seleet individual wavelengtlts, which are

directedtoflnopticalSpaceswitchirrgrrratrix
Alternetivety a wa\ egu
ride-grating demultiplexer
ier could
sould be used
,I
/
'*.
-l

Optical' Commu n rcqtiq! & f'letworking 39s

incontrlng aggregate stream intp individual
-ll[&ihe'
,-t .. Ttie.switch matrix directs the chdrrnels - either to one of the
eightodtput.lines;ortoapanicttlarrbceiverattachedto.the
OXC at output Ports 9'through 12.
o ..,Signals that ar€ genqrated locally 5y a user get connected" :

electrically via the digital cross-connect matrix (DXC) to an

I ' optical transmitter.
,
.' From here they enter thd switcli matrix, lnhich directs them
'....'totheappropriateoutputline..
r .o . The M output lines, each carrying separate wavelehgths, are :

ltiplexer'
fed ihto a wavelength multiplexer; ( ! filuxi and a demd
: is'demux') to form a single aggregate output stream'
? An optical amplifier to boost the signal level for traFsmission :

,.
 -::ffir.'=.{QF.
.l,r+

ree Examination (EC2402-May'Junq 2014 s.

QUESTIoN PAPER CODE ti 2$274

B.E,/B.Tech. DEGREE EXAMINATION, MAY/JUNE'2012.
SEVENTH SEIVTESTER

#,
h**
:
,,.

-
:1.
 ,.='@r.qthr,
',F
s
Cornnrunrcation & ECe402-May-June'201
PART g--f5x 16 = 80 marlc)
--'

l.l'.(a):(i}:..$rawandexplaintheaaceptanaeangleattdnumerical
aperyute of an uptical fiber and derive expressiorrs fbr
botlr.

I l. (a) (ii) A$ber has a coieradiqs of 25 Rln'I, aore retractive index

of I .48 attct relative refractive index difference is 0.0 I . If
the operating wavelength is 0.84 fftrrr, find tlre valueof ,

Deternrine the ,iumUer of guided modes il.n is redueed to

0.003., (8)
'.t...
;
'I .i

il, (b) (i) Draw and exfitain the'refraetive index profile and ray :

aa

transmission in single mode and multimode srep index

fibers and graded index fibers. Write the expressions fer-
the numerical aperture antt number of guided modes for a
Sraded index fiber. (8)
I t. (b) (ii) A stgp index fiber has a core diameter or 7 pm and aore
refractive index of 1.49. Estimate the shortest \\,&velenEh
oflightwhichallowssirrglemodeoperationwhenthe
relati*e refrhctive index difference for tlre fiberi$,l %. ,
(8)
12.{a) Explain the followingwith neceFsary diagram and expiessions
(i) Non linear scattering loss and fiber bend loss. ( l0)'
(ii) Material dispersion io.opiical fiber. iOl

l?, (b) (t) Explain mechanical splices rvith neat diagrams. (8)

t2. (b) (ii) Write a brief note on fiber alignment and joint loss. (s)
I3.'(a) {i} Draw and explain purfaee and edge emitting LEDs. (B)
:- I 3. (a) (ii) Explain any two irtjection laser struetures with neat

diagrami. (8)
.,' (or} .
t3.(b)(i) Explain the operation cf APD with neat diagram. (B)
13. ( b) t'ii) A gilicon p-i+r photodiode ineorporated int.o an optical
reseiver has e quantum effieiency pf 609.t when opereting
at a wavelength of'0.9 mm" The d&rh'e urrent in the device

i
.#
€
 ==------ffi,a-ila,

'&.

C
l
-
.- - BjEjts,lech.DogieeExamlnaHLEClflz:Mayluna2012)
is 3 nA and the load resistance is4 KQ. The incident
. ,,: , optic3l.power.is nW and the pgst derection;ban{width , ,,
, ' ,i ' '
, -100
of the reteiver ls5,MHz, Calgulate the ryet mean square
.........'(rms)shoInoiseandtherma!9oise9urtentsgenemted.

;l4.(a)(i)Discusstlrentlisemddisturbancesaffectingtheoptical
detection systems.
I 4: (a) (ii) Draw and explain tlre operation of high impedance FET ard BIT
plifi€t's. (10)
' (or)

(i) Attenuation measurement using cut back techniques. (S)

(ii) Frequelcydopahmeasuiementbffiberdispersion (8) .

I
: 15. (a) Explain the principle of solitons and discuss the soliton param-

' :,

b (or)
l5.(b).Writeshortnoteswithneqessarydiagramson:.
(i) OpicalCDMA. (8)

t.
 --kca_
"{_feig"
t

e.
'id

Optical ftmm1ln1gq[ (EC2402:@

n CODE :2'0274
B.E./B.Tech. D EGREE EXAMINATISI-: I\{AY/JLINE

ELECTRONICS ANT} COMMUNICATION ENGINEERII\G .4.

EC 24O7IEC 72IIOI44'EC 7L}*OPTICAL

COMMUNICATION AND NETWORKING
(Regulation 200S)
for B.E.
(Common to pTEC 2402 - Optical Communication and Networking
t*t;t*., Sixth Semesta, nrcctronics and iommunication Engineering i'

Time:Thr:g-.n1,rtt,.
" Maximum : 100 marks
Missingdatacouldbbsuitably:assumud..

PART A{I0 x | *:20 marks}

l. The relative refractive index difference (A) for an optical fiber

is lo '. Determine th€ critical, angle at the core cla dding
interface if the core refractive index is 1'46
' '

Ans: Given
A- l%
flr: | '46
-r

q-n2
[=
-,ry"'
0.01 x 1.46 = 1.46- n"

..t
rt}.
ry

: Sln:l- --:---
l '4454
r.46
-'
= sin (0.99)
: 8l .89". -i
,::

:'
,,9

?. A step intlex fiber has a normalised frequency (V) of

26.6 at 1300 nm. If the core radius is 25 W, find the
numerical aPerture. I
 ,)
-* Bf /B:Tech ree Exarnination (EC2402-Ma 2012\
'Ans: Given
v- - 26.6
}"=l300nm
fl - 25 prn.
NA = !

v- fura
. (NA)

T
k
E-.

f,it-

1:
3:
,j

s
 ,.=.ftr:."{ilHilr,
,F
s.

ECa40X-May:June 201?
Photon$ of energy I.53 * are incldent on a
l0-re J
: photodiode which has'F {esponsiViry of 0.qS A/W. If
the optical ,power level is:10 *Wrfind the, photocur-
reRt generatetl.
Ans: (iiven.
 's.
'b

B, E, rB,,Tech Deg ree E xarn i n ation ( EC e403- M ay"! ufi B_20 J_2J

: -r' . . : ,r. .. j PART-BJ5x:'16 = 80 marks) :

llr {a) 0) Draw rind erplaln the tceeptence angle and numerl' '
.fibdr rnd derive expFes-
cel aperturc of an optical

Ans: Acceptance Angle:

I S,ince only rays witt! a sulficibritly shallow grazingangle-(i.e., with
an angle to tlre rttmal gt6ater than Q.) at the core cladding interfade
aretransmittedby1otal}nternalr'ef]ection,it'iselearthatnottllrays-
entering thefiber core will continue to be propagated down its lengtlr.
The geometry concerned rvith launching of light intoan optical
fiber is as shown.in Fig.
. ,,\ *.
1

Loet
By Radia$on

Fig, : Acceptence angla 0. with launehing llght lnto an optical fiber -'
F.
L:,

r. ty A at the critical angle $, within the

k fiLrer at the core cleddinginterface,
j\
:

' It is observed that this ray ente,rs thefrber eore at an angle 0, to

the fiber axis ahd is refracted at the air-core interface befor€ tr&rls-
missiont0thecorecladdingintertbce4tthecriticalangle

angle greater thah 0., wil! be transmitted to the core cl;dding interface

.at an angle lessthan $.' a[d

will not be totally internally reflected.

h
 ' *r.,
.,.l+
B.
\
Optical Communication & Networking (EC24021May:.&ne ?-g'.-}
This is illustrated in the figure, rvhere the inciden! ray B 4t an :

angle greater than 0,, is refracted inttl the cladding and eventually lost
-t

by radiation
a ..'

Thus for rays to he transmitted by total i'ntematr reflection wilhin

the fiber core they must be incident on tlre flber core r,,,'ithin an
,/

acceptance angle defined by the collical half angle 0o.

I ' Henc€ 0, is tlre inaximum angle to the aiis at rvhich light may
../
enter the fiber. in order to be propagated. and is after referred as
acceptance angle for the fiber.
Numerical Aperture
The meridional ray is shbwn in f ig. for a Step Index fiber.
The' tight 'ray enters the fiber c.ore from a medium of refractive .

index n at an angle eo with respect to fiber axis and strikes the corb-
ctaddirrginterfaceat,anormalanglef.'t

Fig.' : Meridional ray optics representation of the propagation mechanism in

an ideal step.index optical $raveguide

reflected. then the meridional ray allow azig-zag ptth along the fiber-
core, passing throu$lr the ax'is of the guide after each reflection.
.FrcrmSnell'slaw'theminimumangle0,,,.ntlratSuppor1total
internal reflection for the nreridional ra) is given by.
llt
sin 0,n.,irr)'nl
=:
 ,. aw is given as
n sin 0o (*",r) = ll, sin 0.
Snell's:!awatPoi,ttB ^^'
nI sin Q. = fr2 sin 90o
n, sin 0, =' rtz
fltL
sin 0. = llr
"t

ll'tL
Q'' = slil-'
.at--+

I
',,
r sin 0u*,ru* t = fr, sin (90" - 0r) (since 0. ='90" -
- nr cos 0.

= lll I - sin2 0,

= fl1 :[tl

v
fi1

ffi
qf
a
t"

=fl u lmax,
I

i Numerical Ape$Lre is defirled os,

NA = tt sin 0o (ma*) =

Numericat Aperture refers to the light collection efficiency of a

given fiber. lt is alsotlte figure of merit of a fibe1. Thus those rays

having entrance angles 0,, I ,o(ma;, are said to be totally internally
 ,:=sF.q.u;}
"l+
g.

I Gornrnufibation & Nahruorki {8C1403-Ma "June 2012

refleCted at the Gone*cladding interface.
Numerical Aperture of Step Index F'ibercl

;VC = ,{ffi
As ntb flz ,

ffC +, 2q (n1

,..s
a

* lt,
ffi
ffi-,,r)
V,'
=frt

th -th
"t '-'t.
where A- '

\ t

Nurnerical Aperture of,step index fibers is giveR as

.4

11.(a)(ii) A fiber hes I cd'rr radius of 25 prryr, cone rcfrnctive in*

dex of 1.48 and relative refractive index difference (A)
is 0.01. It the operating wavelength is 0.84 rl#rr find
: the value of norm&tiued freqiency and the number of
guided modes. Determine the Rumbel,of guided modes
if A ts reduced ]tU 0.003.
LUEtLl {o UrtUlfdr (8)
Ans: Given a = 25 p,m ,;
'-l = -l .48
n. .
I

i
B,E,lB,Teeh Examination Eez{02
I {gne 201 2}
:

, i _+_*.+_

Ln.x 25 x I 0-6 x I .4g.,,8; 0-0 L

= 39.1 1
0.84 x I-0-6

.,

No. of modes le{ = 764.7

765 modes.
'

If A is redueed to 0,003

y?F ?na
x ,-q.€E
t,

2n{25xl0axl.48 2 x 0,03
0"84 x -A
l0*o = 2l.eg
.:/ 'j.

v2
.'. M=--
1
a
=226,6

= 227 modes,

t l;(hXi) Draw and explaln the refractive index profile and rey
in single mode and multimode srep index
'; :lensmission
I fiberc and graded index fiberrs.'Wrlte the expres*ionb
, ' for.th*rnumerienl aperture and numher ui gulded,
!, modes for a g'raded index fiber.
'Ans:
;#;"e ioaexlpfile and rey r*n**ission in sing,- J:i-
.

The
F"- and multimode stqp irrdex nuars;raeaia;il;il;;, is as ihown in
Fig.

c Variations in material eomposition

ofthe eore give rise to Step Index
Fiher and Graded Index Fiher
a) '$tep Index (SI) fihen The refraetive index of core is
uniform rhryushptir and undergo€s un AU.;;;-;;;;;;h;
? core refractive index is
made to vaty fiq a ftrnctiern of raelial distance from the eentre
of the fiber,

v
 ".',/,^ a.
'fi,

qptgisommunicatigl-*,1'lsuglttp,9 (F9?1.0,2:Y!l!YIe ?912,)

j .- -- :.. r- ,9 ' t
e step*incle* and gt.*d index fibersran be further div"d:d
'
into single mode anrJ mtlltirnode fibers'
a) Single Mode Fibers
[tsusta!nsonlyonemodeofpropagation.
-..ri'

of
It.contains several hu,dqeds of modes. A feru
typiqal sizes
.singlernodeandmuitinrodefibersare.given.infig.
I TYPic3l

lndex Profile Fiber cross section and Ray Paths Pimensions

n2l rn1 -+ 12sprm
i2a
t- --J---
-T 8-12pm
- -r-- --J-
(core)

_J . -T
IVlonomode stePindex filter
nrl lnr J 123-4bolrm
tl ----r
(cladding)

I1^- {'I 5o-2oopm

J_e (core)
--T
-J Multimode steP-index fi lter
r 125'140Pm
.v-
'n'! lnr --r- (cladding)
II r=a .
' I . -----
-T 5o-1oopm
\r-I=o - -J (core)
1

-J Multirnode graded'index filter :

step index and sraded

Fig. : comparison.of sins$r
Xltj:#l;iil:lode
Fiber is
ression for nunrerical aperture of Graded Index
given'as I

[-ff ror,=ol
NA(r):{["(r]
-u$1"' = NA(0)
I \o) I

t:0 r>u) I

for

rvhere At{(O) * Axial numet'ical aperture '=

 t=9
; *. , ., ?,[l9t.=n PPge? r*giX.,[ql (E9?{$rr4?Y;r,$1,q3913}
RayT.r.smissionin(iradedIndexFiber'
,' }The reason for less interrnodal distortion irr multimode Graded Index
fibers'than multimode step index fibers is due to radial variation of core
refractive index. It, can be examined by considering tlre light ray'
congruenceWq**ationpathss,lrowrriqFie

, -.. , :. ,,:. , a

f
::
,_,, '
Fis. : The t ive index prof". in a multimode sraded
i*1niJ:?rsmission
Since the index'of refraction is lower at the outer edges of the
core, ligiht rays will travel faster'in this region than in the .entu of the
core where the refractive inilex i'b higher.
.
.t/

This can be seen from the relationship

i

c'
n
rvhere v - speed of iight. n -- refractive index of mediu.m.
Thus the ray cor raracterised by the higher
naracterised hi1 order mode
will tend to travel further than the fundamental ray congruence,, but at a
faster rate. I

will thereby tepd to keep up witlr tlre lower

The higher order mode
order mode, which jn turn*educes the s.pread in the modal delay, (or)
reduces the dispersion in the fiber.

Considering thq ray theory, the raIS travelling close to the fiber
axis have Shorter paths when compared with rays,which travel into the
outer regions of the core. However, the near axial rays are transmitted
through a region of higher refractive index and therefbre travel with a
lower velocity than the more eitreme rays. This compgnsates for itre
shorter path lengtlrs arid reduces dispersion in the fiber.
' A similar situation exists for skew rays which follow longer helical
paths as shorvn in Fig,.

C
 Communrebtion & Networ*i

,n q:, *n*i| t$e i sati

speeds, thus siv i ng the sarn* pg:ha3 i s1
:f le.9al
index
31'
preifile
l"tence *ultimode graded index fibers with parabstic
i ndex
eopes have largei tran$m i is i on, bandwidth s than the multi mode "step
fibdrsi
: Consequently, ahhough, they' qre nst eapable of bandlftlis
aftainable wiifr single mode fibers, suqh Rlultimod: glt{* inl{ex fibers
have ihe advantugi of large eore diarneters eouple$ with bandwidths
suitable for long distanse aommunication. --

Fig. phowq the mode number plane ittusffating theScrd3 boundary

and the guided fiber modes. Modes are said to be four fuld degenerate.

t rtr il
l.{t"'
ii.:.t.')
.. guided fiber
Flg. : The mode riimber illu$retlng the mode boundary and the
modes
Defining the mode foundary as the function
i-
mr =f(L),then the
total number of euided'modes M is glven by
,

fo,r*

=4 I t$\ dL
 i*,.
B.E./B.Tech E 02-May-June 2A12)

: 3' intCorresPondilg P four modes '

' ' ' ' ttl

,c

r in the fflor
element'of unit area
an element mode Plarie:

M=la-l'(rrr
s \a+//
ka)'L

t'
. 14s4
=

.11, (b) (ii) A step'in:dex filrer has a cofe 1fiame!". 9-r 7 Y and
: core refractive index of t ,:,49.'Estimite 'the shortest
wavelength of tigtrtlwhich allows single mode opera-
tion when fhe relative refractive index difference for
the fiber is lo/o.

fr' = l '49
?\ = 0.0 I
Y - 2.405 (Siugle mode operation)

2ran1v2A
.J

. r
^
fv
-
V

ytrx3.5ilO-!x1.49 2 x 0.01

= I.92 x 10:6 m (or) I .92 pm.

 -,:EhFr_,-]G*.,
't

s.
"l*

tical Communication & Networking (EC2402-May-June ZAn a

- 12. (a) Explain the following with necessary diagram and )

expressrons
.a

-
(il Non linear scattering loss and fi.ber Qiend loss. (r0)
(ii) Material dispersion in optical fiber. (6)
Ans: (i) Non-linear Scattering loss .

Optical waveguides donot al',ways behave as cotnpletely linear

channels whose increase in output optical power is directly proportional
to input optical' power. Several nbn-linear effects occur. which in the
case of scattering cause disproportional attenuation, usually at high optical
power levels.
This gauses the opticatpower from one mode to be transferred in
either the forward or backward direction to the same or other modes at
a different frequency.
.

The types of non-linear scattering lstses:

I ) Stimulated Brillouin Scattering (SBS).
2) Stimulal.gd Raman Scdttering (SRS):
These mechanisms in fact give optical g?in but with a shift in
frequency, this contributing to attenuation for light transmission at a
specific wavelength.

i Stimulated Brillouin Sea,ftering

(SBS-)

It rnay be negarded as the modulation of light through thermal

. molecular vibrations within the fiber. 3s

\v
Itis an interaciion between light wavqs and the vibratiohal modes'
of silica molecules. The scattered light appears as upper and lower
sidebands which are separate(from the incident light by the modulation
frequency.
The incident photbns in this scatteringprocess produces a pHonon
of accoustic irequency as well as a scattered phonon
:

scattering angle because the frequency of soundwave varies with

shif is maximum in the backward dirgction reducing'

The frequency
to zero int he forward clirection making this process as a mainly'
backrvard process.
 - r=?{F,..+eg,
'.F
E.

B.E./B.Tech
It is significant above a threshold
The threshold power P ois given by,

PB.= 4.4 x l0-'. * tr' oon . lr' . watts.

d - fiber :ore diameter in: Fffi,

I - operating wavelength in pm.
0," - attenuation in dB/km.
V - Source bandwidth iR GHz. '

This expression allows the determination ofthreshold optigal power

which must be launched into a single mode optical fiber before SBS
. occurs.

qxcdpt that a high frequency optical photon '

It is sirnilar to SBS
rather thar an accoustic phonon is generated in the scattering process.
SRS can occur in:both forward and backward directions in an optical
frber,, and may an optical power threshold. of upto three orders of
magnitude higher than the Brillouin threshold in a particular fiber.,
The threshold opical power for SRS P- in a long single mode filer
' Po,= 5,9 r t0'2 aF'l ao" watts. : .L

. --._----.
I hrs ls due to:
| ) Higher dopant concentration;
. , 2) Larger scattering loss due to greater compositional fluctuations.

. In addition, multimode fibersare subject to.higher ordrir,mode

losses owing to perturbations at the core-cladding interface.

Fiber Bending Losses

Radiative losses occur wftenever an optical, fiber undergoes
a bend of finite radius of curvature. :

Two type of bends are: ,'

, a) Macroscopic bends Having radii that &re large com-
pared to fiber diarneter.
b) Random Microscopic bends Arise when fibers are
incorporated into cables: i

F
 t

Optical Communication &, Nelworking (EC2402-May-June 2912) l5

 q '-=4rr'.'"IQQG'
'.
g.
^B

B.E.lB. Tech' Degree Exaniination EC2402-Ma -Juneffi1 t.*,'.

I Externat Force
I
I

,1,

Fig.iacompressiblejacketextrudedoverafiberreducesmicrobending

' Wlren the torces are aqplied. thejacklt s"ilt be defor{ned but the
fiber'will tend to stay relatit'ely straight as irt figure.
I higher order modes are Lror,lnd less tightly ,.:. th...fiber core
:-
than the lower order modes, the l-righer order nrodes will radiate out of
a'

Thus the total number of rno{m th4t can be suppotted by a berrt

_
. . a l ,!t
F

,.. 'fhe effective number of modes &';r, that are gltided by .a

curvedmultintodefiberofradiuS.a'isgivenaS
f f- -i - .rzlrll
ll
=N* l'-si3 l+.["'=l
where ry. - total number of modes in a straight fiber given as:"

mber'of'modes in a

D
core radius 'rr'-and outer raditrs
{p. u multirnodei Cl
CI fiber, having
h
'b' ths,rnicrobinding loss u* of aiacketed fiber.is reduced from thal
of an unjacketed fiber by a,factor. .

., -'2
€
:

(
,(b)+ :i,Er I '
F(uo,,)= ll*n'\'l:
t
\(/'' Ei ) I

E,It and E - \'oung's'moduli of fiber and iacket respectively':

..-.,.
 =ffi!:ru'llE.
'i

s.
"!{

Optical Communication & Networking ( EC2402-M,ay-J une 201 2) t7

(ii) Material Dispersion in Optical fiber
index of retiaction varies as a function
'
of lv'avelength.
This is shown in F,ig. fbr qilica.

1.54

152
lndex 1-50
of
Refraction, 1:48

I 1.46

1.44

1 .42

1 4 oo
?nirr",lfn,r,?,,L,

As a consequence. since the group velociry V,,of amode is a furrction

of the index of refi'action n. the iorious spectral to*ponents of a gi;;;
mode will travel at different speeds, depending on the wavelength. -
"Io
caliulate material induced dispersion, consider a plane warie
propagating in an infin:itely'extended dielectric, medium that has'a
refractive index n(1.) equal to that of'the fiber core. The propagation
constant b is thus given as.

A
a- Znn(i.)
P- I
L
The group delay r oilr
vr
'da do d It_tzncl
W.K.T. l/=
$
1-
since
,l?r= ,tt"L Il
,dF
dX da
dg dx
f ri
= 2ncl
--.- -+
L i.2 j
i

I
 's.

l8 J.

d
= i;rl",
i
zn l-' rls)
r..-l

J dx T.T l"
^

l, -;r-J
-l4rl
'Lt'
b,
r-
o- ?nf
,L
a4

"mal ,
vg

=
L( Ldnl )
flrI
-lc(
For a source ryith rms spectral width o^ and a mean wave-
lterial dispersion. o,,, ryA)-,
be obtained from the expansion of equation in Taylor's seriesabor *t'I
w-,here

. O=
-tn rv
d?" dz?\
 --
'I

g.
"!{

::

alttated b""'' considering the

dependence of r,,, ol"l i,.'

{:,;, =cL I ir, --ji{n1- #?r

t"dtl I

4\, L,/i ;^ J

,:rr:!jg
c 'dx?
OMA=$.. f l^'#\ =or,'Ll.D*u, (2')l

where ,D,u
o, (A) isthe Material Dispersion.

1.9., Dn',o,
..rr

l
l-l
a'nl
(or)
lar'I
which is expressed irr ps ntn-t km-t -

12. (b) (i) Explain mechanical splices with neaf diagrams. (8)
Ans: Mechanical SPIices
lnical spticing, in which the fiberi are held'in aligrrrnent b.v..
som€ meclrartical llleans, l*ay be achieved by various methods
lncluding
placed (groove
the use of V grooves into which the buffed fibers are
splices) or ttre use qf tubes aro-und the fiber ends (tube splice-s)

In the y.grooye splice technique, the prepared fiber eQds are first
L ^
^l^ -- -.-
butted together in a V-groove as shown in Fig. .,They are then bonded
together with an adhseive or are held in place by t1.1:.t of a:over
plate. The V-shaped channel can be either a Brg?:*d siljcol: .Plastic,
cerarnicn or metal substrate. The splice'loss in this rnetltod depends
strongiy on the fiber size (outside climensiotts and core diameter variatidns)
ani eecentricity.'(The position of'the core relative to the centre of,the
1

liber)

-i

vi
 *
'r f+
g'

B.E./B.Gch Degree Examination {EC?402ffi une 2012)

Fibers Butted
a.

Together

V.Groove Substrate

Flat

irhrrrra cnJinac

:ally, perform lateral, loirgitudirral and arr$ular

alignrnenr. It splices nlultfrnode frbers to give losses in the s1m:range
as commercial fusion splices, but ntuch less equipment and skill are
geeded. The splice mechanisrn is basically a tube rnade up of ann elastic
..c

material . The central hole diameter is slightll' smallerthan that of'1he

fiber to be spliced and is tapered on each end for.easy fiber insertion.
 "---! -EIE-
'. f+
.tr g.
,/

tical Communication &:Networktn ( EC2402-May-J u ne 201 2)

Capillary Tube Size

Less than Fiber Rad

-\

Fiber Being lnserted ,- lnserted Fiber

Tape!red
Opening

Fig. : Schematjc of an Elastic-Tube Splice

' When a fiber is inserted, it expands the hole diameter when it is .. .

, inserted so that the elastic material exerts a symmetrical force on the. r

fiber. Tlris symmetry feature atlows accurate and automatic align-
rhents ol'axes of the two fibers to be.joined. A'wide range'of : fiber
' diameters can be inserted into the elastic tube. Thus the fibers to be
: . spliced do not have to be equal in diametet since each fiber moves
:

l1 into position independently relative to the tube axis.

F=
f
12. (b) (iilWrite a brief note on fibef alignment and joiirt toss.
(s)
Ans: Flven when the two jointed fiber ends ar.e srnooth and perpendicular
to the fiber axes, and the rwo fiber a.res are perfectl! aligmed in a small
portiorr of the lighi may be ieflected back into the transmitting fiber
carisirig attenuation at the joint. This phenornenon, known as Fresnel
F refle.ction. Themagnitude ofthis partial reflection ofthe light transmitted
r
t1
a

E.
through the interface may'be estirnated usingthe classical Fresirel formula
for{iglit of nonnat incideiiCe and is given by
?

h ,
r
&
B
h

r' /
i6i .

r2
r = Ilnr-n)-
l_ll.l
a
\q+n)
where r is the fractioh of the light reflected at a single interface, 4, is
the refractive index of the fiber core and n is the refractive index of the
rnedium between the twojointed fibers. However. in order todetermine
'
the alnount of light reflected at a fiber joint. Fresnel reflection at both
tibers interfaces must be taken into account. The loss in decibels is due
lo Fresnel reflection at a single interl'ace.is given b1.':
 .

'' l{
q.
1) rraminatio
Lil- B.E.IB Tech oeUree
# 2-May-June zafi)
Hence.usingtherelationslripsgiveninEqtiai
ion at a fiber-tiber joint.
 . i ! s1:'-!--------..*n!I!h:

. *
g,E b Tech Degree Examination (8C2402-May-June 2012) .
Hence. using the relationshipi gi
the optical attenuation due'to Fresnel reflection at alfiber-tlher joipt.
ItiSapparentthatFresnelreflectionmaygiveasignificantlossata
fiber joint even when all other aspects of ,i. corl,r-..tion ur*'il;i
However, the effect of Fresrtel refleciion at a fiber-fiber connection
can be reducep to a very low level through the use ofan iidex rnatchipg
fluid in the gap betwe.n, h. jointed fiblrs. When the inrJex ,o*t.hinl
fluid has the sarne refractive indexas the fibercore, losses due to Fresn.t
refl ection are eradicated.
There are inherent connection problems when joining fibers. -
o
Differelt core andloi cladding diameters.
. I
. Pipt:i:,rr",rerical apertures and or relative refiactive
,

. Fiber faults.
The loss6.s caused by'the above factors together with those of
Fresnel reflection are usually referred to as
intrinsicjoint losses. Examples
of possible rnisalignment between coupred compaiib.leoptical fibers dre
illuitrated in Fig. It is appargnt that misarignment may occur in three
' dimensions, the separation between the fibers (rongitudinal misalignment)
F . in the offlbt perpendicular to the fib", ax-es (lateraVraoLuaxiai
rnisalignment) and the angl.e between "or"the core
axes (angular
r
t optical losses resutting from these three types of misalignment
depend upoir,the fiber.type, core diameter and the distribution of the
optica! power'between the propagating models. Exarnples pf ine
measured optical losses due to the various types of rnisalignment aie
shown in Fig. shows the'tttenuation characteristics for both longitudinal
and lateralmisalignment of a 50mm core diameter graded index fiber. It
may be observed that the lateral misarignment gives significantly greater
lossEs per unit displacement than the longitudinal misalignment; For
instqnce in this case a lateral displaoement
of r 0 mm gives abour I dB
inssition loss'whereas a simitar longitudinar displacement gives an
insertion loss ofaround 0. I dB.
..-

h
 '.1 : .- ' *
! ... . .,
-June 2012) {t
Optical Communic

Fig,:,fiiernree pdssibte tvier of,misallgnment atwhicl ,ty o|.I yltn

ioiiiing cirmpatibte. oqlcal ri\r;
{a)'longitud.inal: misatignrnent, tb} Iateral
rRrsalignment, (e) angular m isalignment
Fig. (b) shorvs the attenuation characteristic for'the angular
*irtuignmlnt:j two lultilode :t.q. indel fiber1 yTI .nuTtrical
lpertures of 0.22 and 0.3.
An insertion.lijss ofaround I lPr::l,iined
with angular misalignment of 4 and 5 for the 0 .22NA and 0.3^IA fibers
respectiv*ty. It may also be observed in Fig. (U) tlyt the effect of an .
index *ut hi6g fluid in the fiber gap eauses increased losses with angular
o i.

-.
_ Therefore
it is clear that relativqly small levels of latllal or
1nf
angular misalignment can cause signific.ant attenuation at a fiberjoint.

.:,3.0
lnsertion
Loss (dB) .

Longitudinal

20 25 30 3s

Displacement (rrm)

(a)
 .:ffi-
[F.
's.

B. E iB, Te'ch Degree Examin ation {EC 2402- May'l une!0 I 2)

40 Irrdex Matched
gap nA - o.32
30

?.o

lnsertion
Loss (dB)
1.O
Arr Gap
NA-O3
o

2345078910
M isalignment Angle (Degrees)

{b}
Figs.: lnsertion loss characteristics of jointed optical fibers with various
Upes oi misalignment (a) Jnsertion loss duq to lateral a1d longitudinal
rnisalignrnent for a 50mm core diameter graded index fiber {b) lnsertion loss
r- due to, angular misa!:l1**nt for ioints in two'multimo.de step index fibers with.

E
r-
,.
(i)
:

13., (a) Draw and explain surface and edge emitting LEDs.
!

Ans: r In this configuration, the plane of the active light emiuing

ll I
' . . t' t ' '
F. region is orierrted perpendicularly to the fiber axis.as in thq ,

substrate of the devicef ihto which a fiber is cemented in

\
order to accepithe emitted light. The circular active area

'. The emission pattern is said to be isotropic with l20" half

power bepm width. This pattern is generally called-,lambeffian
pattqm.
.a'
 ,..-q**r-xjQp..
t tlq

B,

Optical Communicatiorr & Networkrng (EC2402-May-June z}y

Fiber

Circular
Etghed Well
rialieation

r Double-Hetero-
I iunctton Layers

Active Region

Fig, : Surface-emitting LED

? In this pattern the source is equally bright when viewed tiom

emitting surface.

It consists of an active junction region, which is the source of ,

./ as shown in Fig. 3.4. The
incotrerent tight and hvo guiding layers -.-"

refractive index of the guiding lay,ep have refractive index lesser than
that ofthe Active layer b; higlter'thail the refractive index of sulrounding
-medium. This struci'rr* fonng waveguide channel that directs theoptical
:OT
radiation toward the fiber iore.
To match the fiber 6re
diameters (50- 100 mm), the contact stripes
for the edge emitter are 50-70 ffiffi, the lengths of the active regions
usually range from 100- I 50 rnm.
\

o In the plane parallel to the junctioilo there is no waveguide

is rlmbert'ian wltn a harr power-
;lT,T'# '':'l;3":*u*
par

o In the ptane perpendiculat to the junctiono the lralf power

bearn width 0l r pet'penrlic'ular has made as small as 25" to
i0" by a proper choice of the \vaveguide thickness"
 '.. Fr
E

't

EEIB. T.ctr Deg ree Exami n ation-( EC 2402-May-J un e zo 1 a

The ernission pattenr of edge ernittei
compared to that of the surtace emitter.

Stripe Contact '

[Defines Active Arca]

,t
Metalization

[For Electric C6ntact]

Si02lsotatton Layer.

Substrate
f.

Metabolization

[For Electric Contact]

Heat Sink
lncoherent Opthat
!"

l
r
:

F
{
t Fig. : Edge.emitting LED

13. (a) (ii) 'Explain any twd injection laser structurCs with neat
diagrams. (B)

A basic requirement for efficient.operation

.t of laser diodes is rhat
in addition to transverse optical and carrier confinement between
heterojunction iayers, the current flow must be restricted lateral[1 to a
narrow stripe along the length of a LASER. :

. The goals of some novel rnethods are:

, l. limiting the number of lateral ntodes so that lasing is confined
- t
I
,'.€,
to a single filament.
?2.-stabilising the lateral gain. .
ir.

.*i3.ensUring;relativelyr1*in,e,t,oIdcurrent.
Fig. shows basic optical confinement rnethods for copfining laser .

f,
 Z)
l-Communication & Networking (EC2a02-May-June ?9t
l'tigh.Refrac{ive Lor-Refractive
lndex Regiuts

0Pticall
'
{'w'1
/rL\
Oplical lntensity I
htensity
ooo
-10 0 10

.?0 0 20

(a) (b)

H[h-Rcfractive

Lateral
Far-Fbld
Patterns

Fig. :,Three fundamental strugiures for confininq optical !n.tne

vtfay'gs q:Tl
wavegui e, (c) negative
direction. (a) gain induceu gujdg, (b) posttivg indu

.ln 1[e fiLrljtruc-turg,.e electrode stripe (less than 8 mm

1u:ro:
rvide) runs alonfi the length of the diode.
i all tl:" lengtll
t:'."* the
along
The injection of electrorls d holes into,the,device i
of the diode alters tlre refractive index ofthe active
layel dirlctlv bel3rn'
;.;ripe The profile of thpse injected carriers creates a rveak. complex
\\'aveguicle that confines the light laterally"
 ,.'-rEr:g':q[r..
't"!t

28
fnis Upe ofdevice is c,ornmonly refg.rred as =Gain Guided Laser.
''
Alttrough,these lasers can emit opti0alpowa:s,excee$it g *,G0tnW
tfi€y have strong instabilities and can have highly astigmatic, tvt,o peaked

31

k
 Optical Communicatron & Networking. (EC2402-May-June 2012) , 29
\.
one etches a Rarrow mesa stripe (l - 2 mm) in double heterostructure
material. The tnesa is then embeddect in high resistivity lattice-matched
n-type material with an appropriate bartd gup artd low reference index.
.. This materia! is Ga Al As in 800:900 mm lasers with a CaAs acti$e
layel,and is IhP tor 1300 - !600- Rm lasers with an InGaAsP active"
layer.,T[:is'configuration thus stiongly traps generation light in a lateral
waveguide.
p.(laAl,A.s
crmfining rr-lnP

Ncgitive
tontrct
!lrlrfs\r

Fig. (a) Short Wavelength (S00 -.900 nm) GaAlAs BH Laser Diode,
(b) Long Wavelengths (1300-1600 nm) lnGaAsP (BH) laser diode'':-

,:
13. (b) (i) Explain the operation.of APD with neat diagram. (S)

Irr this diode, the primary photo cunent is internally multiplied beforq
it enters the input circuitry of amplifler c'ircqit. This increases the re.celver
'
sensitivity. sinee the photocurrent is multiplied before encpuntering the
thennal noise associated with the receiver circuit. b r
In order forcarrier multiplication to take'place, thd photo generated
carriers must traverse a region whene a very high electric field.is pres€flt,
 _ r1Tj6!'r'.",ruB.
, ,l+
B,

B.E./B;Rch Examination (EC2402- ne,20:12

+
n
&"
{F, WM
t P

k t
F Depletion
i(n)
-

F Region

+
P

through avalanche photodiode structure and the"electric fields

in the depletion and multiptication regions

gain enough energy so that it ionizes tlre bound electrons in the valence
band upon colliding with thern. This carrier multiplication mechanism is
called impact ionization. ,

The newly created carriers are also accelerated by the high electric
field, thus gaining enough energy to cause further impact ionization.
This phenornenon is known as avalanche effect.
When a low reverse bias voltage is applied. most of the potential
drop is across the pn' junction, The depletion layer widens with increasing
bias untilacertain vottage is reached at which the peakelectric field at
the pn* junction ,is about 5- I 0% percent belovv that needed to cause
avalanche breakdown. Afthis point, the depletion layer just "reaches
through,'totlrenear:lyintrinsicpregion.
Normally, RAPD is operated in.the fully depletion mode. Light enters
the devices through the p- region and is absorbed in the p region which
acts as the collection region for the photon generated carriers.
Upon being absorbed, photons gives up its energy thereby creating
electron hole pairs in the p region, which are then separated by high
electric field in the p region. The photo generated electrons drift through
the pregion in the pn* junctioh, where ahigh elec'tric field exists. It is in
thishighfieldregiorrcarriermultiplicationtakesptace.
 '?{F -ill$q
..9
B

F
lq_ , ____r , B.E./BiTech
' '
Degree Examination
, , -
(EC2402-May-June
\- a ,
2012)
-.-r

+
n

If WM
,'P
F t
F Depleti0n
F.
i(n) Region

+
P
t

k"
Fig. :,Reach through avalanche photodiode structure and the"electric fields
m in the depletion and multiptication regions
fiL,

gain enough energy ro that it ionize$ tlre bound electrons in the valence
band upon colliding with them. This carrier multiplication mechanism is
called impact ionization.
The newly created carriers are also accelerated by the high electric
field, thus gaining enough energy to cause further impact ionization.
This phenomenotr is known as avalanche effect.
When a low reverse bias voltage is applied. most of the potential
drop is across the pn- junction, The depletion layer widens with increasing
bias until acertain voltage is reached at which the peakelectric field at
the pn* junction ,is about 5- I 0% percent belovv that needed to cause
avalanche breakdtr**n. Afthis point, the depletion just u'reaches
layer
through,,tot|renearlyintrinsicpregion.
Normally; RAPD is operated in,the fully depletion mode. Light enters
the devices through the p'region and is absorbed in the p region which
acts as the collectiorr region for the photon generated camiers.
Upon being absorbodo photons gives up its energy thereby creating
electron hole pairs in the p region, which are then separated by high
electric field in the p region. The photo generated electrons drift through
the p region in the pn+.iunction, rvhere a high electric tield exists. It is in
thishighfieldregioncarriermultiplicationtakesptace.
Optical Communication & Networ:king (EC2402- 2012
lonization Rate:
The average numLrer of electron hole pairs created b!' a carrier
per.unil distalce travelled is,cal,led ionization rate. Most ntaterials exhibit
different eleitron ionization rates (u) and hole ionization rates (P).
-flrerefore
the ratio t * $la of the two ionizatiol rates is a
-n:reasure.of the photodetector perftlrmance. '

The multiplication Ir4 fora,ll carriers generated in photo diode is

defined by
. M
rvt = Ir,/ln
r[,rr rfl --
I* I

ionization,

The perfonpance of dn APD is charaeterized by its respolrsitivity

R , which is givan as

: '] R^nu= (X) r=rou

.:
whgre Ro - unity gain responsitivity.

t3.(bxu) A silieon p-i,:n photodiode incorporatcd into ari optical

! . receivgl hag a qyantum efficiency of 69q/" whep
', ' in the Oivipe is 3 n^A qnd the load resistapce is 4 KQ..
'The incidensoptical power is 200 nW and the post
detection bandwidth of the regeiyeris 5MHz. Calcutirte
' the root me4p square (rmq) shot noise apd thermal
. noise currents generated. (S)

-\
7::0,9x l0*'m B=5xlffHz
t1,,=3xl0-eA, r ' T='293k
' 4,t=4kg2- t ^
t':'':
 ._,ry_:,rullE
'J,q
S'
*
it

') ^r
)/. B.E;lB.Toch D Examination (EC2402-May-June 20 1 2)
Ans:
lvlotr
tD 0.6x 1.6 * l0-19 x 0.9x l0-6 x 200 * l0-.9
ro-
trc 6.624*10-34x3*108
t-
I
,,: 86.9 nA

.i,,,, : 2q 1,, B F.f W : 7q IpB, whEre M' a1d F(M) - l.

^rF
:Z x 1.6 x lO"jo x 86-9 x lO-n x5 x I0"'
E
.i ,'= | .3904 x o
l0-l: A:
^1

.ir,', : 2q Ir) BArf (FM) =Zq I D'B; rvhere M und F(M):l.

F
E.

*a
a
,-
<i,.:>
'lt
: 4.8 x 10-21 Al .

)"i ' R7
"L 4000

-t
2.A217 x ,10 -r7
.g

<i,j>' = 2.,0217 x l0-r7 A2. :

a.

(a) (i) Digcuss'the.noise and disturbances affecting the optical

detection systems. (6)
The errors in $e detection system can arise from various
noises,and {isturbances, associated with the signal detec-
tion system'as shown in Fig..
Neiise is defined as unwanted components of an electrical
signal that tend to disturb the transmission and processing of
signal in a physical system.. The noise is caused by
spontaneous fluctuations of current can voltage in electrical
circuits.
 .-=_<a- -r@[i!..
..:{
'8,
Optical Communication & Networking (EC2402-May-June 201 2) 33

External to the Svstem Internal to the System

Eq uiprnent gcnerated Noisel Quantum noise/slrot noise
:

Surfirce leakage cunent noise

'l'hermal
noise

The two mos[ common samples of these spontaneous fluctua'

tions are shot noise and thermal noise. q

. Shot noise occurs in electronic devices due the discretehature

of current flow in the device. '.

t It is caused by the randorn amival rate of photons on the

photo detector.

t Thennal noise arises from the random motion ofelectrons in

a conductor.

Photodetector
Gain (M)

p a Photon + Bulk Dark Current o Thermal . Amplifier

detection o Surface Noise Noise
quantum Leakage Current
.

I
h' noise 1 Statistical Galn
B.
:- [Poisson Fluctuation [For
&
E
r,
fluctuationl avalanche
F photodiodesl
.Y
Ii
Fig. : Noise Sources and Disturbances in the optical putse detection mecha-
il
t In Avalanche photodiodes an additional shot noise arises frorn
statistical nature of multiplication process.
o Additional photo-detector noises come from the dark current
and leakage current. These are independent of photodir;de
a

illumination. Therefore it is rnade very smalt in relation r.r ith

the other noise currents.
 '&
' . ' l'':;"@
:34 . B.E,IB Tech Degree Examination (EC240&May-{ute ZgiC1 '
 's.

3,5
Optical Cornrnu n icatior
d irt a given'time slot, rnost of the
pulse ener5/ will ar.rive in the col'responding time slot. in the
receiver as shown Fig.
in
Due 10 pu'lse Spread'ing induced:-by flrc fiber some of the
" fran.sinitt engrS, will progessive.fy spread' into'neighbouring
tirne slotS As the pulse propagaibs aliing tiie fiben

.o T6e presence of this enersr' in adacent time slots results in

lth
an i nterferin g s igntl, he nce e term intersymbot i nterferen ce .
tn F ig. 4,.6,thq fraction ofenergy reinaining in the appropriate
tirne slot is designed by y so that 1 - y is the fraction of
energy that has spread into adjacent time slots.

(l}
B
.J
,B' C
(,)
a
E
B
E-

Fig. : Pulse spreading in an optlcal signal Sd leads to in:ssryrnbol interfereT.

explain the opemtion of high impeda.nce FET

a t

14. (a) (ii) Draw -!

"pd
Ans: Pre-arnplifiers used in-optical fiber communication rec'eivers
can be classified into three broarJ categories:

3) Trans Impedance Pre-amplifiers-

The Low Impedance (LZ) prearnp is the-most straight for-
1'
rvard, but not necessarily the optimufil pt€-amp design.
ln,this design, a photodiode operates into a low impedance
.,] amplifier. (eg. 50 O). Here,, a bias or a load :Tislor Ru is u1d to
match the amplitier impedaRce. The value of Ro; in conjunction with
r,*.
,"5

,-t'

.6-':l
€
 ."84s-'a@E:
,!q
"
B'

JO B.E./B"Trech' Degree Examin ation (EC2402-May-J$ e 2A12

fiei input capacitance is such that the pre-amp, bandwidth is
the amplifier
equal or greater than the.signal b''andrvidth,"
1

' This limits their use to special short distance applicatiolrs where
-
In the HiEh Impedance (H.Z)- Pre-amplifier design, the.goal
is to reduce all sotrrces of noise to tlre absolute minimum and maxi-
mum. This is accomplished by decreasing the input capacitdnce
through ttre selection of low capacitance high frequency devies, by
selecting a detector with low dark currents and by minimizing the
tlrerma[noisecontributedbytlrebiasingresiStorS.
The thermal nois6 can be reduced by using a high impedance
amplif,*. [Example: tsJT or FET] together with a large photodetector
bins resistor Rn. Hence called'a high impedance pre-amplifier. .

!The, Tranb hnpqffance Pre-amplifier desigt , largely over-

comes the drawbacks of high irnpedance preamp.
This is done by utilising a low noise high irnpedance preamplifier
with a negative feedback resist* Rrrvith an equivalent thermal noise

The amplifier has an input equivalent series voltage noise source

eo(t),,an equilalint shunt current noise ldr) and an input impedance
given by pprallel combination of R, and C;.

High hnpedance Amplifters

lET
k.
A number of different FETis-can be used for front end receiver'
a

h The circuit of a simple FET amplifier is as shown Fig.

f
 .FigI Simple high impedance preamplifier design using a FET

The principal noise sources are thermal noise associated with

the'FET channel conductanee, thermal noise from the load or feed-
back resistor and shot noise arising from gate leakage current. A
fourth noise source is FET I noise. f
Since the amplifier input resistance is very large, the input curent
noise spectral density S, is
.-.
S,r, I'ET = __:_ b* I gate
+,Zql gate
.Ra
_. Zq I
gute

Bate
FEr. i;r
In an FET, the thermal noise of ttre conducting, chirnnel resis-'
tanceisclraracterised.bythetransconductanceg'. rr/nr

ThevoltagenQisespdctra[dbnsityis.'-:,..
'r 4K"IT
\=
AD-

-Eo'
'6m
f - FET channel noise factor which is a constant that accounts for
thermal noise and gate inddced noise plus the correlation between
these two noises.
The thermal noise two characteristic s W at the-equaliser output
is then

4KBT ,4Kn7rl , ,(znc)'+ '--"-:

tlz+i -,-:.i
KaTr ,
r-.:*-{fi
R6 E,uRi I t q ) sm I$ r
t

#l'!qtgaie

#
 :

38.,'
where C'-C,
"
+(a +C*a
.A.e.q.recrroegleeFxaminariontecza0z;ay-Juffffi$ffiln
+C*s
The lf- noise corner frequency{ is defined as the freriuenrqy,. ,'
atwhicht6noisa whi-ch dominates tirirgr noise at low fiequencils
has a l//power spectrurn, bocomes equal to the high irequency
channel noise described by I-.
The high irnpedence amplifier design technique yields low noise,
results in a low dynarnic range.

High Impedance Bipolar Transistor Amplilier's i

The circuit of a simple bipolar - - ---- em
r - --- grounded ----:' -"'r""''
-- itter transistor ampr ifi er
:
is as shown in Fig. . .

V*t(0
Photodiode

Couplirg
Capacitor

'Fig : $imple high impedance preamplifier design uqing a bipolar transistor

l.

KaT
R=_
trt'n--
. qh;

For a low noise design. R-, and R. are chosen to be greater than
R,, so that Ra=lRin .

The spectral density (in A:lHz) of the input noise cun'ent results
frorn the shot noise of the base curreht.
 Erwl*::QE:
'9.
J+
..

optical'iomtnunrca@ (8C2402-May-J une 201 2)

2K BT
S, = 2qIas='

.:. The sPqctral height (in tnlHz) of noise voltage source IS

q=
. \JF
t o
6m ,;
' qI -p
,{
[" 'Collector Current.
Substituting the above, wd get

',yy ri2KBiL[* . - {2nc)z

= *. #),, rl
l"t

If the photodetector bias resisto; Ro is greater than amplifier

.au R^ then R =
Ru = R;r -so th4t,
o'hfl

resistance

,
\tr/_
TL *,, B -2=' W J

As in the case with a high impedence FET prearnplifier the

impedance loading the photodetector integrates the detector output signat.
irnp"nr*t. for this, the amplified signal is differentiated in the
equalising filter.
:'
Trans Impedance l*Ptr$er
Although a lrigh impedance design produces the loivest-noise
amplifie+it has fwo limitations.
1) for broadband application's. equalisation is requi,red.
has a lirnited dynamic range.

An alternative design is trans impedance amplifier show'n in

-,.
Frg ':

is basically a higher gain - high irnpedance u*piifi., wi$

feedback provided to the amplifier.through R, This designer yields both
lownoiseand.1argedynamicrange.]

*€
 .. r.Bj:.
E:
''
'l+
g.

40 B,E./B.Tech -June 241.2)

we make
the riestriction that both have the same transfer function
a

For the trans-irnpedance amplifier tlre thermal noise characteristics

:

W,,rat the equaliser output is therefore found simply by replacing Rb

with R;.

I rltt'I
;
I r .t
7-:i-
t

R' RRfRoRbRf
- -

k
a
 - r=+g.,..]W8,
'r t+
g.

,
.1

"&' ECz40z-Mav-June 2012)

tical Communicati0n Ne

For the trans-impedance amplitier the transfer tunction Ht'z(./')

IS i-

;,
1
I

I +'.yr? TI RCf I A

'
whiclr yields a bandwidth of By7ti= ;*.
4ftc'-
'fhis makes the
which is A times of the high i,npedance design.
equalisation task simpler in frequency amplifler case.
:--
Benefits of Trans-irnpedance Amplifier:
t) Wider dynamisrarrge, cbmpared-with the high impedartce amplifier
design.

2) No equalization is fequired' .a

I resistance is small, so that ampl'ifier is lesso susceptible to

pickup noise, crosstalk and electrcrmagnetic interference (EMI).

4) The transfer characteristic of the amplifier is actually its trans-

impedance which is the feedback resistor. Therefore the tmns-
impedance amplifier is very easily controlled and stable.
.r a I r I r
^t

S) Although fansimpedance design, is less sensitive than high impedange :

amplifier [... WirrW,rr], this diffbrence is usually only about 2.3 dB

for rtrost practical wideband designs.

14. (b) (i) Explain, the following measuroments

.r-t'

(i) Attenuation mea$u rement using cut back

techniques. (8)

(ii) Frequency domain measurement of fiben

dispersion. G)

*?
h
 42 EC2402-Ma n6'201
Ans: (i)

Order
ll
I
-;,j

Sorting lViewing
0ptics
Filter

Fig. : A tYPical exqerimelt anangement for the measurement of spectral

loss

The most commonly used technique to determine the total fibre

attenuation is cut-back (or) differential meihod.
I ) It consists ofa white light source, usually xenon arc lamp (or) tungsten
halogen.

2) Achopper is used which chopsthe white light to few hundred heftz.

ck-in amplifier at receiver side to perform phase-
sensitive detection. Phase-sensitie detector can .**.t
signal from a
extreme ly' noise env ironnrent.

4) The chopped light is then fed through a monochrornator which ,s*s a

prism (or;.diffttction gratin-g arrangement to select required
band of
wavelength where attenuation is to be measured
5) Hence the light is fi{tered before it is fbcused on to rhe fibre by means
of a microscope objective lens.
6) A beam splitter may be incorporatetl before rhe fibre to provide light
for viewing objects and a reference signal to compenr.t. for out[ut
v' I
power fluctuation. '
 '.J+
'9,

*
43
tieal Copmunr-c9$on A
qn a multimode
it gives
fiber, the launch optics shoutd be arranged in such a way that
steady state mode distribution.

s1 ror.lhis purpose a dummy oo: is used, u1o-1lso a'mode scrambling

deviceisattachedwithinthefirstmetreoffibre.
'.

9) The fibre is also pttt through a cladding miOistrrn{1*:'*h lonsists

of S-shaped groove cut in the Teflon and is filled wittr glyeerine.
be used at the frble,end to remove optical
lg)-Mbde stripp.er can also
po*"rthat hur sgatrered from the core t:,o.claddrng t:r
, 1*.n :pri:ul
powgr at receiving end is beingdetected usually by a PIN or Avalangh,e

Photodiode.

1)The phorodetector is usually indep matched to ou!-p1t .n1 of.filre

usrng epoxy "resrn (or) any index matching cell for obtaining
a'

lZ) Hence the final electrical gutput from photodetector is fed to a

Iock-irramplifierandtheoutputisrecorded.

Thg cut back technique involves taking a set of

ry*"ti:u::tt-
ment'atoutput over the required wavelength in a long fiber. This fibre
is generally uncabled. The fibre is',cut back to a point !-metres
from

inputend.The[aunchconditionshastob.emai11a':*:..
given
The optical auenuation per qnit length a dB fur,fhe fiber is
AS

l0 r--
Iog'o
Psz !

t"-t" e;
Lr -+ original length of fibre
L., -+ cut-back length of fibre
Po, -+ output optical power at:t,:lial.Yavelength of fiber
Po. -+ output powef at cut-back fiber lengths.
'
.

Lr and L, are measured in km.

 28 B.E.1B.Tod DegreeExamination (EC242-May-June z}fil '
;'
I
A'ltho$gh,these-lasers,c*n.ernit te@ryq:sexceed$rgls0nW,
,they*rave:stronginst*bilities,andcan]ravefr.ighty amigrna*ic,two pe$cd
beams'&s $hown in ,figure {a).
. More
,stable stiuctures use the.configurations shov*n in {igure (b)

' : Here the dielectiic waveguide structtres are f,abricated in thelaterd

these streq$uresraop&d the:lateral modeq in the laser.i '

I ,, ..

Thus t&ese
.
devices:are
:.
called Index Guided Laseils.
1
': ,
.:
{f a partifular:index guidi:d laser sgpportB only the,flndamental
tr&nsverse rnode and the fundamental longitudina! modg, it isknoivn as
a Single Mode Laser. Such a deviee emits a single, well collimated
bearn oflightthdhasan intensity profite which is a belt shaped Gaussian
curye.
: Index G$idd tasers can have either.positive index or negative
i nd ex wave-con fin,ing struct urss.
]central
,''[n a'positive-,inder wav€rr,id€,'t*te , rqgiorr tras ajhigher
refer,e.rlce index than the quter regions, I ..1

T'hus all ofthe guided lighr is reflected at the die{ectrip boundary,

just as it is at the core-cladding interface in an optical fibre. .

In a negatfue*index weveguide, th€ central region of the activ

has a lower reference index tftrow the outer regions.
"tayer
' 'Af tlre dieleailc boundaries,,patt ofthe figlft i's re{Iected alrd thg
rest is refracted into fie surrounding material and is thus lost. This
radiation {oss appears in the far-fieH radiation pattern as narrow side
lobesto the main beam as shown in,Fig. 3',2,+ 1c7 -Sinc€-th€ ftlndar$emel
mode in this device has less radiatiod loss than any other mode, it is the
first to lase. The positive index taser is the more popular of these two
structures.
-
Index guided lasers can be made using any one of the four
fundamental structures.'l'hese arq
l) buried heierostruciure. 2) a selectiVely dift'used structure.
3) varyin$thictness structure, I 4) benl layerconfiguration"

31
 .-q*sr".ruE
'"^!{

Sr

I Communicatton & Networking (EC2402'May'June ztUL ,29

(BH) laser as in Fig' ''
'
one eteSes a frarrow mesa stripe ( I - 2 mrn) in double heterostructure
'fhe
material. tnesa is then embedded in high resistivity lattice-matched
n-type material wltlr an appropriate band gap and low referencg index.
.'_.ffrii material is Ga Al As in 800-900 mnr lasers with a GaAs active
' 1300 r 1600 nm lasbrs with att InGaAsP aptive
eonfigur:ation thus strongly: traps generation light in a lateral
llfll.l:
waveguide. r

lhsni?s ctrnnct
rnrt hort sink

oqilttci
(a)

-
(a) Short Wavdlength (800 900 nm} GaAlAs BH Laser Diode,
+ Fig.
(b) t ong Wavelengtlis (1300,1600 Em) lnGaAsP (BH) laser diode
l',-

13. (b) (i) Explain the operation of APD with neat diagranl. (S)

In this diode, the:primary photo curnent is internallymultiplied before

}'-
Fi-
it enters the input circuitry of amplifier circuit. This increases the re.ceiver
sensitivity. since the photocurrent is multiplied before encountering the
i-

thermal noise associated with the receiver circuit.

.a

, In order for canier rnultiplication to take placen the Pl,otq generated'

carriers must traverse a region:where a very high electric fibld i3 fresent.
 i5rijff'*t-
f.
,.9
g.

t
:*
B.E./B,Tech Exarnination' (8C2402-May-June 20 1 2

Electric Field

WM
P Avalanche

t
Depletion
i(n) Region

+
-__l__
P

F is' : Reach'n'iJlLi:rll.;l; erectric rierds

:H'lt;ilH:','liTT.'rilfJhe
-,
' ' In this high field region, a photo generated electron or the hole cap
gain enough energy so that it ionizes the bound electrons in the valence
band upon colliding rvith them. This carrier multiplication mechanism is
'
called irnpact ionization. .

The newly created carriers are also accelerated by the high electric
field, thus gaining enough energy to cause further impact ionization.
This phenomenon is known as avalanche effect
When a low reverse bias voltage is applied. most of the potential
drop is across the pn- junction. The.depletion layer widens with increising
bias untilacertain t:oltage is reached at which the peakelectric field at
the pn* junCtiorf is 'about g;1 0% percent' below that needed to ciuse
avalanchb breakdoun. At'this point,.the depletion layer just "reaches

Normally, RAPD is operated in the fully depletion mode. Light enrer$

the devices through the p; region and is absorbed in the p ,egiln which
acts as the colledtion region for the photon geJlerated carriers:
Upon being absorbbd, photons gives up its energy thereby creating
electron hole pairs in the p region. which are then separated by high
electric field in the p region. The photo generated electrons drift through
the p region in tlre pn* junctioq; where a high electric tield exists., lt is in
_

this high field region carrier multiplication takes place.

Optical ConrmUnicatron'&, N
Ionization Rate:-
Thp average number of electron hole pairs created b,'- a carrier
per unir disrance tra'vdled iscalled ionization Most.rnaterials ibit
1ut:. 1f
diff"erent electron iqnization rates (u) and hole ionization rate,s (P)'

Tlrerefore the ratio ft = F/a of the two ionizatiort rates is a

meaSureofthephotodetectorperforlnance.
Multiplication Factor

defined by
.

tr*- average value of total,multiplied output current after impact '

ionization.

The performance of an APD is characterized by'its responsitiviry-

ft^,r,,whic,h is givelps
.,
.

(wl , _h,M
Ia*,,/
whgre Ro - unity gain: respopsit iv it-v.

IS:(bXii) , A silicotr p-i-n photodiode,incorporatcd into an optical

operating at a wavetrength of 0.9 mtlt. The darh current

in the device is 3 nA and the load-+esistance is 4 KQ.
: The inpiden$*optical power*,is.200 ,*rillJfiflil:
detection bandwidth of the receiver is S I!---
the.root mean square (rms) shot noise and thermal
oise currents generate{. (8)
\\
M:0.6 P,,: 200 x l0 ' W'

1 :0.9x 10 m o

tIr, :1x l0-eA

t*t"t ''
, i
R, = !, kg.
.

i
t..
 .:----frg_ ,.i]iQGr

'9.
',.8

. $. 't
1,rr
J/, B.E.lB.Tech Degree Examination (8C240@ l

0.6x 1.6 x l0.19 x 0,.9-x I'0-6, ZbO* l0'9

T
D:
.u

'-
t

I r'- 86.9 nA

.i,r,r :2q Ip. B lr? F(M) : IiB; where M and F(M) = l.

x.'*l0'n x 5 x l0('
x I10-Io,A
I

1.3904
. .):.

-zqlD'8, rvhere M and F(M):I,

x lO"n y 5 x 't05

x293'
Rt 4000

i -''
= ).0217 10

.i,?r=/.02t7 x l0-l t7
A2.

14. (a) (i) Discuss the noise and disturbances affecting the optical

Ans: The errors

a
iqhe detection system caR arise from various
noises arld disturbances, associated with thg signal detec-
-tion
t r
-a

system as shown in Fig..

Noise is define-d as unwanted components of an electrical

signal that tend to disturb the transmission and processing of
signal in a physical system, The noise is caused by
spontaneous fluctuations of current can voltage' in electrical

b
 _..',sarr,.,'%f,G.

a
qp
"F

Optical Communication & Networki (EC2402-Ma ne 201 2)

.External to the System lgternal to the System

Equ i prnent generated N oi sel

,/,
Dark current noise
Surfhce leakase cunent noise
.

The two most.oommon samples of these spclntaneous fluctua-

tions are shot noise and thermal noise.
t Shot noise occuis in electEorric devices due the discretenature
€
of current flow in the device.
. It is caused by the random arrival 'rate of photons on the

o Thennal noise arises from the randorn motion ofelectrons in

l
Photon
t,
F*-

& Stream Blas Resistor

s.
F
l-
AA/V{ AA'+
r' rhv
1
r
b.-
I
l,

F""

I
L

[r
F
H
i*; t
t
& . P.hoton + Bulk Da* cunent . Thermal o Amplifter '
-S*-,. detection . Surface Noise Noise

: .

[Poisson Ftuctuation [For

fluctuationl avalanche

\
Fig, : Noise Sources and Disturbance.s in the optical puls/detection.mecha.

. In Avalafiche photodiodes an additional shot nolse arises fi'orn

statistical nature of multlplication process.
o Aclditional photo-detector noises come from the dark current
, and leakage current. These are independent of photodiode
illtrmination. T'herefore it is.nrade very small in relatiorr r.r ith
the other noise currents.
 B.E.lB.Tech Examination (EC2402-May.'Ju ne lQl!
tor load resistor R, and the
amplifier ilput resistance R,. Since this noise is of Gaussian
in nature, iqcan be eaSily treated by standard techniques.
\
T[e analysis'of the noises and the resulting errer prohabilities
associated with primaqy photo current generation and ,

avalanche multiplication ane complicated since neither ofthese

processes is Gaussian,

o The primary photd curent generated by"the photo diode iS p

time varying poisson-process resulting from the random arrival
rate of photons at the detector.

lfthe detector is illurninated by an optical signat P(t) then the

averagenumberofelectronandholepairsgeneratedNina
' time interval t is given as .
Y

Tl f, Ig
,\' = *Jrtrl dt = hv

where E is'is the enefgy receiver in the'time interval t,

F .

:
q is the detector quantum efficiency
I

The actual number of etectron and hole pairs 'n' that sLre
generates $errerated fluctuates from average according to

- .;-- tl -Nir
, . -ne
^r pr(n)=,N
_,

B intervaht.
H
r"*

For a detector with avalanche gain !M', the excess noise

:i*

K
&,*

p x -+ ranges between 0 to l'depending on the material.

F. A fufther error source is attributed to intersyrnbol interference

::'
h
(lSl),which results from pulse spreadirrg ih the opti:al fiber.
[:
l-.
i-

$
 '.:-*Er,1i]Qf.,
'.. Jx

€.
,t
: /
35
ffieal Cornmunicatisn & Networking.
@e.iitr6nsmitted,irtagiroentimes1ot,mostoft1re
pulse enerry witl arrive in'the corresponding tirne slot, in the
receiver as shown in Fig..
.t

Due to. pu:lse spreading induced by tlte fi'ber'some of the

transm i ued ene rgy'wil{ progres s i ve ly spread' i nto ne i gh bo u ri n g
tirne slots as the pulse propagates along tire fibe-r.

o ' The presence of this energy- in adjacent time slots results in

q-

an'interfering signal, hence the term intersyrybol interfer€nce.

ln Fig. 4.6,the fraction ofenerg), remaining'in the appropriate
- time slot is designed by y so that 1 - y is the fraction of
energy that has spread into adjacent time slots.

l-^!

t4.(a)(ii)DnawandexplaintheoperationofhighimpeoanoeFET

Ans: r,.-#1 ffiffiT:fi;, fu'Jr communication receivers

: '
t) ttieh Impedance Pre-amplifiers .

The Low Impedance (LZl preamp is ttre most straight for-

ward, but not necessarily the optimum pre-Amp dqsign.
ln this design, a photodiode ciperates into a low impedance
amplifier. (eg. 50 O). Here. a bias or a load resistor Rois used to
match the amplifier impedance, The value of Ro, in conjunction with
 i.

36 B E.lB Teeh:Degree Exam i n ation.' ( EC2402- May..l$le 2AI 2\

theamplifier input'capacitance is such that the pre-amp, bandwidth is
equal oi gpster than the signal bandrvidth.

high sensitivity is not a major concern.

In the HBh lmpedance (Iflrz) Pre-amplifier design. the goai
is to reduce all sources of noise to the'a'bsolute minirnum and maxi-
mum. This is accornplished by decreasing the input'capacitdnce
through the selection of low capacitarrce high frequency devies, by
selecting a detector with low dark currents and 6y minimizing the
thermal noise contributed by the biasing resistors.
The thermal noise can be reduced by using ir high impedance
amplifier [Example: BJT or FET] together with a large photodetector
bins resistor ^Rn. Flence called a high impedalce pre-&tnplifier.
The Trans knpedange Pre-amplifier design, largety over-
comes the drawbackq of high impedance preamp.
This is done by utilising a low noise high irnpedance preamplifier
with a negative feedback resistbr Rrrvith an equivalent thermal noise
current +(t) shunting the input as in Fig..

'.af

Fig. : Equivalent Circuit of Transimpedance Receiver Design

The amplifier has an input equivalent series voltage noisesource

-
er(t\, an equivalent shunt current noise i,(r) and an input impedance

A number of different FET's can be us€d for front end receiver

designs. ,

The circuit of a simple FET.amplifier is as shown Fi-u.

{
['
 '.F
g.

tical Communicati0n & Networkin (EC2402-Ma ne 201 2

V*t(0

,
,l

The principal noise sourceis are thermal noise associated with

the FET chann.i .ooductance, thermal noise from the load or fbed-
back resistor and shot noise arising from gate leakage eurrert!. A
fourth noise source is FET lf
noise.
ier input resistance is very large, the input curent

4 Krr
I, I.ET = Ro
.1

= 2q I eutr
,1,

where I*.,. - gale leakage curfent of FET.

In an FET. the thermal noise of the conducting, channel resis-
tance is charackrised by the transconductance g*. '!

The voltage noise spectral density is

\rE- a
Es
affi

f - FET channel noise fuctor which is a constant that accounts for \

thermal.iroiseandgateinducednoiseplusthecorrelationbetween
these two'noises"
The theimal noise two characteristics IF at the equaliset output
is then

Pf=
h[,'t*u
 -.'Fr-aiFr-_
:.-
+
.
*
l,.E

'B.E,E}.Tech
-June,2Afi
where C=6'-a +C'd*Cra *(.**
tC'a +C; +('^^ :

'',, The
, tU *
I ne llf . nolse
noise corner ti"equelcy{
frequency/is defined
d, as the frequency
at which lfnoise. which
dominates'the
the FET noise at low frequencies
f."o,r**rciei
has a llf power Spectrufft, becomes equal to the
high frequer.y
channel noise described by I-.
The high irnpedenoe amplifier design technique yields low noise,
results in a low dynarnic range.
.v

HighImpedanceBipolarTdansistorAmpliIier's
The circuit of a simple bipolar grounded emitter trbnsistor amplifier
is as shgwn in Fig. ., .c

;'&? Fig : Simple high impedance preampllfier design using a bipolar transistor

The input resistance of a BJT is given by

k
Pti4?r,:$uT
-
{itt" gIan
-"Elr*
Ji.

K where Iu* - baie bias current. ',.

r'
F
ry*
a

For a low noise design. Rr and

& ** ts be greater than
"
chosen
F
R.tn so that Ro =,R n .

' The spectral density (in A:/Hz) of the input noise cun.ent results
 - r*drjdr. *..
EE,
'' Jt
eF

, (8C2402"l\IaY-Junb 2012 39
Oftiml,Comm

t,u 2 -*- --u--

2q I gg
Rru

height (in WlHz)of noise v'oltage source is

2K BT
\\'E- o
6m.
;

81, B

lc Collector Culrent.
Substituting the aboveo w€ get
a

\
2 R,r l,
T-r-..--.-tt't
(}nc)z
+--._;-
w =?zKBr[[*. RD pna,)- r; ?4-]

I I,f the photodetector bias resistor ftn is greater than ampl''ifier

resistance ^Ro then R -- Ro = Rin so that'

rl lv
o
r'.frr rt*WR,, 'rJ
w - ry[ &, B '!,.-
^ t., *l o , 1 l1,rr,.\! |

It "tn l'
'

Tl gru
.,E

'A, in the case with a high impedencg. tlt preamplifier the

output signal'
impedance loading the photodetector integratesth* 1.1::tor
in the
Again to comp*,r*te for this, the amplified signal is differentiated
equalising filter.

Trans Impedance, AmPlifier

amplifier,ithastrvolimitations..'':
l) fol broadband application's, equalisation is required

,_.'. *
An alter.irative design is tran.s impedance amplifier shown in
Frg;
This is basically a higher gain - high impedance amplifier:irl
y'ields both
feedback provided to thla*plifier through R, This designer
 ;F
B.

40"
gnt, we make
the restriction that'bsth have the same transfbr function

Hou,U)t H ot.{.).
For the trans-irnpedancsamplifier the thermal noisq characteristics
is-therefore found simply by replacing Rb
,rat the edualiser output
with R;.
Rofi
(t R6+ft7
Rl^=

' ,Also lll

Rffi Rf Ra Rb Rf

T,4K"T
. ,a'

,'
where y*r- High impedarce A*rrp,lifier, noise c'haracteristic gir:lt ly
either rquuiion for FET designs or by equaiion (4.9) fior the bipoldr
transistor case. rs

Wb not compare the bandwidth of the,l*,o designs. The transfer

functionof.n.on-feed,backa'rnpIifieris(inV/A)isgive!aS

-H(f)-' ' 44-

AR

wnere R .Ro Rb
*frequency independent gain of amplifier yielding
C - C, + C' 4 and A -,
 - iEi4S, *l-
f
.rF
i;+ g.
F'

tical Gommunication & Netvuorking (EG2tl02-May'June 2012)

H0z-May-June 2012) _ 4l
th. t rrn r- i m p ed a n c e a m p I i ti e r t h e t ra n s fe r ftt n ct i o n H 77(.1'\
ls =;r'
I

,l

t+ jTnRCf lA

which yields a bandwidth a,f Br, = i::- .

which is A times of the high irnpedanog design. t

fhis makes the
' tl
.equalisationtasksirn.plerin.frequencyamplifiercasel
.

Benefits of Trans-impedance Amptifier: '

i) WiAer dynarnic rangen compared *ith the high impedance arnplifier

design.

Zi No equalization is required. \
3) Output resistance is small, so that amplifier is less, susceptible to
pickup noise, crosstalk and eleetromagnetic interference (EMI).

4) The transfer characteristic of the amplifier is actually its trans-

impedance which is' the feedback resistor. Therefo-re the trans-
impedance amplifier is very easily controlled and stable. '

5) Although transippe$ance design is less sensitive than higlr impedance

amplifier [.". WrrrW*,rJ, this difference is usually only about 2-3 dB
for most practical wideband designs.
t'f
.

14. (b) (i) Exptain the following meaquremerh'" a .

(i) Attenuation measurement using cut back

(8)

(ii) Freq[ency domaip measurement of fiber

dispersion.
'a
(8)

F
 -.}!!t*r;1'
"i
G.

'r'r
tr
'B

42 a:f,q;Fgh
.t

Mode Scrambler

,'
,
Fig. : A typical experimlnt anangement for thb peas.upment of spestral lose
in optieal fibers using cutback technlque

The most commonly used technique to determine the total fibre

attenuatiorrisbut.back(or)differentiglmethod.
I ) ltconsists of a white light source, usually xenon &rc lamp (or) tungsten
[ralo$en.

2) A chopper is used which chops the white light to few hundred hertz.
\
3) This enables the ldck-in amplifrlr at receiver side to perform phase-
sensitive detection. Phasp-sensitie detectorcan extract signal from a
extreme ly noise env irorrment.

4) Theohopped lightisthen fed through amonochromatarwiich usesa

prism (or) difrhaction grating arrangement to silect required band of
wavelength where attenuation is to be.measurEd. j
5) Hence the liglrt is filtered before it is focused on to the fibre by means
of a microscope objective lens.

. 6) A beam splitter may be incorporated before the fibre to provide light

.. for viewing objects anda reference signal to compensate for output,
power fluctuation.
g ''

g.
^l{

ONcalCoriimunication & Networking (EG.240?:May-Ju.ne

'
412)4.3
, ...
7) tf ttre rneasurement of.attenuationthas to be done on a rnlmmooe
, frtqr,,the,launch optics should be arranged in such a way that it gives
stea{Y state mode distribution'
I l).rnrriis'purpose a dumrny fi-bre is used, and,also a'rnode scramblrrrg
t-d- within ihe frrst mare of fibrE. ' I ,

9),The frb,re is also put throu$r.a,oladding mode strirpe,r which consiss

'
', gf S-rtraped,groove cul in the Teflon and.is filted w,irh glycerine. '
,
l0) Mode stripper. can also be used at the fibre ena to 1m11e optical
. pow€r.that has&enscattered fromthe core into cla{ding. The qptical
power at receiving end is beirig dotected usually by a PIN or Avalanche
Phdtodiode.

ll)The,sotodetEctot is usually index matohed to output snd 6f fibre

using qpoxy resin (or) any index matching cell for obtaining
'reprgt*uciHerestr{ts., ' --'
, ,.'.,',,,'
12) Hence the final electrical output from photode.tector is fed to a
lock-irt amplifier and the outp.ut is recorded'
' of pqw€f r{leasure-
_, The ctrt ba+ ftrhniquririnvoJves taking a set

ment at output over the pquired wavelength in 1 long fiber. This fibre
is generAlly uncabled. The fibre is cut back to a point 2-metres from
input end. The iaunch conditions has to be maintained'
The bptica,l attenu+tion per rurit lgngt{r a dts fof tl}e fiber is given
AS
...
D
t0
-- losrb P
\-h _r.,
r 0t

Ll -+ original length offibre

L., + cut-back length of fibre

Po, -+ output optical por,t

re1lat original wavelength of fiber
_
Po= + output power at cut-back fiber'{enEhS.
Ll and L, are measursd in km.
 B,E./B.Tech Examination (E
Equation ('1) may'be written as

, _ cd*= q_4 to8rof,

V, -t output voltage .eading from original fibre length
V, + output voltage reading from cut'back fibre length.

Drawback
l),Destructive teehnique. ' '

2) Although it is srlitablg for laboratory measurement, it is far from i.de+! .

for attenuation measurements. ,

. .il

ilOXiil'Frcquency domain measunement ofrliber dispercion;

. . Frequency domain,rneasurement is peferred method:for

r€sponse r{or) and bandwidth
. .) oltainins
of
bybana
fryuency
multirnode optieal fibers.
Variabh Avalanche
Diaphragm

Specfum Analyser

Fig. : Exparimental setup rfor making fiher dlspenion'mea$urements in the

frequency domain using a pulsed laser source

t
 .F
g.

whiclr takes.the Fourier transform of the pulse in the time

domain and henge display its frequency components.

o The base band frequenc)- response for the fiber,is,:

Po (ro)
I{al) =
\
4(c,t)
" where Po(co)
\r'
is the spectrum at the fiber output.
P,(to) is the spectrum at thefiber input.

Swept Frequency Measurement Method:

o In this method involes launching a modulated gptical signal
at ditTerent selected frequencies using a sweep oscillator.

Fig. : Block schematic showinSan experimental arrangement for the swept

rresueniv meas u rem ent mffij;rtffit tffi is persion m eas u re m entg i h

o The signal enetgy is very nirrow frequency band in the

f:',[;1,::::il'$,5J;::[:T;'ffi:,ffi ::i#:ll
o Optical source is an injection laser which may directly

o A spectruln analysef may be used in order to obtain a

continuous display of the swept frequencysignat:
:
 B.E./B.Tech Examination une 2A12
The spectruln analyser provides no information onthe phase
ofthereceivedsignaIandavectorvoltmeteroranetwork
analyzer can,'b* employed to give both the phase and the
freq uency i nformation.

An electr:ical or optical reference channel is connected

.-*
between the oscillator and the meter.
.*}
r.l
o When an optical signalo which is sinusoidally modulated in
power with frequency {,, is transmitted through a single-
mode fiber of length L, hlnce the modulation envelope is

-8 L
ffi
4!rb
irS;
'' *!.
;
vg '8
-=r
rl
1

,wllere
I/n is the group velo9ity. .; ? ,

\._
B.
a A.deIayofone*taulationperiodrof}5.-eorrespondsto
I
of 2n, then the sinusoidal modulation i.s phase
a phase shift
shifted iq the fiber by an angle'$,.
2n, sL
,A
.
.tU-m
tm I
zm
=2nf
.' ..rm
r
g
L
a

I
Therefore f* =
.Tm

The specific group delay is given by ts =&

15. (a) Explain the principte of solitons and discuss the soliton
parameters with nece$sary expressiorts and. diagrams.
. (16)
Ans: o Gioup Velocity Dispersion (G\{D) causes most pulses
in as thgy throu$h an optical .

lime
ffl"aden lroo"Bfe I

o 'solitons' are puhs thattraJel along the fiber without change

in shape or amplitude ol velocity.
 ' ..tg
g.

I NeUqrking Gc?403',$gy{ll$,?,@
t Cqmrninicatioh --_-'-- **r---;;
t!t-lq-:t
1I
r Soliton, takes'advantage of noR-linear effbcts in siliea,
 -.-_r@-iUnL,
't

g.
''l{

;t

E
ar
o
.6
a-5

.E nl
(I) v.r
{x
>,
G'
5
o
t,.
- *1,

v
1.0 .

Relative Time

Fig. I Charaqerlstlcs of a high intornlsty shaply peaked:sollton pufee

. . To derive the evqlution ofthe putse shap required for sodium

mansmission, bn* nqed t_o consider the. Noft'l-inear
 - -'!i.;dr,, *.
S.,
.,.^E
qF

'l *i Ou l02u
";,'+ N2 u -.i(cr l?)u
*,=;
dz ZAf lu12

.whereu(,,,r/isthepulseenvelopefunction.

cr, is the of energy gain per unit length, with

loeffic.ient
negative values of a representing energy loss. ,
,

o Special soliton units tc

I elimindte
----. - - scaling constants.
l

-,_- three terms in the R.H.S. of equation (5.50)

The
F"
?
. ThA first term represents GVD effects of the fiber.
i'

I The second rofl-linearterm'denotes the fact that the refractive

index of the fiber depends on the light intensity.
,

'r' Through the self-modulation process. this phyqical

* phenomenon broadens the frequency, spectrum of a pdlse.

. The third term reprgsents the effects of energy loss or gain.-

-laa^

' | .

:The solution to equation (l) for the fundamental' solution is

given by

{ f) = sech (f) exP (izl1) '

tt(s"
where sec,h (r) iS that hyperbolic socant function. This is a bell-
shaped pulse as in FiB.:

1.0

0.8

0.6

0.?

:V 0
-2 -1.0' 0 I 2'
, l.lormalized fime ,'

Fig. :,The hyperbdic'secant function used forsolitqn puises

 'r fl
E

50 B.E /B.Tech' (E9?{Qtlrtay:;qne ?fi?)

The phase term exp (jzlT) in the above equation has no
influence on the shape ofthe pulses, the soliton is independent
ofzatrdhen'ceisnon-dispersiveintlretimedomain..
a When examining the NLS equatioo, it is found that the first-
order effects of the dispersive and non-linear terms are just
,'
complementary phase shifts.
For a pulss given by equation, these phale shifts 3re:

'
for the non:linear process and forthe dispersion effect. i1..,
' | -/r'
u-l
( t o?ulr,=[!-secJvv'*ol)a,
dLairp =l ;
\ )u,Otz )*'.-Lz
Fig, shows plots of these terms and their surn which is a,
constant. Upon integration, the sum yie[ds a cornmon phdse'
t'
shift of zl2 for the entire,pulse..,
Since sqch a phase shift changes neither the temporal nor
the spect al shape of a pulse, the soliton remains completely
non-dispersive.in both the temporal and frequency domains.
t

1'
(E
o
o
o
o 0,5
:o t
o-
o 0
,.8
.G
b
Q)
tr
-0.5
4-2024
Normalized Time
,I

Fig, : Digpersion and non.finear phase shifts o a soliton pulse

+.
 ,:,*.P-.'a-%F
s
'.F

iml Communioation & Networkin

Soliton Perimetens

r lhe Full-Width HalflMaximurn (FWHM) ofa pulse is defined

as tlre full width of the pglse at its half-rnaximum power
t

I 'FaFialtt

The FWHM (ofthe fundamental soliton pulse in nonnalized

'.t

where4i'thebasicnormalizedtimeunit.Thisyields
,l

TT-
,
:
r
2cosh-' JZ I .1627
F
t
[_
r
s
i
r."

F
t
L'

r, .
f.

F
ts-
H
'The nornglized distance parameter (also called dispersion
si' _
L.. -

k
V:
I"-
teristic length for the effects of thd
r

r.

F .\
I
ki't
E

f"
lrrn, a measune of the period of a soliton. It is given by
..'
H
E-

F\
&,r
n*
&
&
tt: ,

ffi
ffi
qiEP
+#*
;dnii.
{::tEr "
idp6.

&-
- i*:;.--

il':
.:
\ryhere c is the spqed of light I A
'.',

-.,..
D is ths'disperion ofthe fiber.
| :- *^ --^l :- l-.-
i is
,-Lo,rn IS measured
TTIE*{SLIIE(I [n km.
III I(III.

e The parameter P,ninis the soliton peak power is given by

Y

}
 - ig*4r*r,,.8ffiffi;.,.
.

s
^."!{

. 52 B:E./B.TedrDeEnoElqmilaqn GC2402;[rray..,uno ZOt2)

4t r ' Lt.lozr)2 Aun?u3 D

\

,D_G_=rLLL|-J_+ r.o* ,ql..,2t l.l

N,,^- 2nn2 ilzc, fi
wherc l, is the effective area of the. fiber cbre.

a, is the non-lincar int.r.ity dep"ndent refractive index io-efiicient- ,;

Fi.

F
. - for trf > l', the solition pulse exper/ences periodic chaiiges in
itsshape apd spegtnrm as,it propagtltes through the fiber.
. . it resumes its initialshape at multiple distance of the soliton
'.*{..
t. "7r
periodr L1,rr,r,,rl = Lo,r,
;

{ecrrartt
1.0'

0s
{ TJTo

I
,flE
!
:l l\
,10 -0,5 0,5 1,0 I

Fig. : Definitlcin of the half,maxlmum soliton wifih in tErms of normalised time

.. if unlE :

':
.Y4
o : tf TBis ihe width of.bit slot; then we can relate tlre bit rate B
:
,l

I---=---
- .76
.- rJ 2Ss% 2So&
-,."t r.
; ..
i

where the faetor 2So =

tietween
ffi is the normaliied sepffration
:

ne i gh bori ng so I itons.
. ,'

{
I
 -r:lQ*r
.rF
g.

gptioalComrnunication & lilehrorkihg (EC2402-May-June 2012) 53

. o For solitons that are initially in phase and separated by
-i
j

TI

' :.i .t '. l

. . o. The mutual interactive forbe between in-ptrase solitons thus

results in periodic attraction. collapse and repulsion.
. The interaction distance is,. .

- L=(lL.
-t -- -.loP -L .exo(&)
-1reriod -':-\\-o' .

. I This interaction distance, and the ratio lJLu,,ndetermine the '

maximuh bit rate allowable in soliton ty.tt oit.
' . I The criterion Q Z ,
: .loP'2> L..where
!'
Z,is
.l
the total transmission
.:: distance.'
. .. The design condition
:- {l Ldi"p>> Lr

)roo - m;:.1
8,-4..[,.sor"
_t
I g 1 ltf n___r\_u/ o62ro

the above equations hows tlre effects on the bandwidth ^B or

, ,..',
the total transmission distanpe trr,for selected'values of ,S,.
,.i.
15. (b) Write short"nstes witb necessary diagnms on:
(i) Optical CDMA. (S)
(ii) WDM and EDFA system performance. (8) i
Ans: (i)

.by.assignrngl"uoiqu: co{e tg each

l.*t:
? To communicate witll andther node, users imprint their agreed
, upon code onto the,data. The receiver can thdn decode thg.
bit stream by locking onto the same code sequence.
r , The principleof qptical CDN{A is based on'sirqead-spectrum
techniques.
 E .:F;{'ia..-fq*,

'.. F
e.
t
54 B.E./B.Tech Degree Examinatitin (8G2402-May-June z}nl
.; The concent is to spread the energy of the optical signal
. over a frequency band that is much wider than the minimum
bandwidth requifled to send the information.
Spreading'is done by a code that is indepen&nt of the signal

On optical encoder is used to map each bit of information

into a h i gh-rate (longercode-length) optical'sequenie.
O The symbol is the spreading codes are called chips.
o The enerry density ofthe transmitted waveform is distributed
more or less un,iformly over the entire spread-spectrum
-/ bandwidth.
I
,l

n
I
. . The set'ofoptical sequences becomes a setof unique 'address
codes or signature sequences' the-individual network users.

F
Data signal
b

Signatu"
Sequence
t'
E

Transmit.d
\ Srgnat "Tuur.---dY*u-:'T-

' Tlre signature sequence .oituinisix chips. When the data signal
contains a I data bit, the six-chip sequence is transmitted, ro chips are
sent for a 0 data bit.
,{r
Tlme-domain optical CDMA allows a number of users to access a
. ,....-i 1

networksirnultaneouslythroughtheuSeofaeommonwavelength.

been examined; In synchronous accessing schemes follow rigorous'

transmission schedules, they produce more successful transmission
p rrouglrputs) than asynchronous methods when network access
is random and collisions between users can occur.
k
b
I
:'

h
 1a

-I-
iC2402-May-June2012) ,. 55

Stital or
Electrical
- Sources

fi

i
i, :.

.N xN
Optical
Star Couphr

rt-.

Fig. : Example of an optlcal CDftIA netulork based on using a coded sequonce ,

of pulso

['-
F,.
r
F

B
!
$
i
L

::
56 " ''B.f.n.runD€stwExaminationGCZfO2'l,tirJune2012)
.'- - : $
': o - For a rEceiver to
'be
able :to.
distinguish the proper addrebs " j
. correctlyl it is neeessary to maxirn,ize:the, autocorrelation
function and minimize the cross-correlation funetion.
r Prime-Sequencg codes and optical orthgonal codes (OOCs) '

. An OOC system the number of simultaneous

,' \ users N is

r tr-r I
u"l *-f: :1"- l-

where'F is the length of the code sequence.

-'
K.is the weight or the number of ones in this sequence.

If the N transmitters in a WDM liirk such,as in Fig. with at bit rate

B, throughB* opei'ate respectively then the total bandrvidth is
9f
N

: i=l
/ "'
When all bit rates are equal,'then tgtal bandwidtll is N times the
channelbandwidthofasinglechanneIlink1
Forexampleifthebandwidthof1clrannel*2.5cb/s' t

: then the total bandwidth for I channels = 20 Cbls. and for 40 channels
= 100 Cbls.
Tetal capicity'of the WDM link depends on:

the available
transmission window.
Standard wavelength spacing - ,100 CHz by
ITU-T recommendation G67 ?r
Optical po\iler requirements for a specific BER.
At the outputs of demultiplixer, system parameters that need to be
corrsidered include 'Z
t+
l) Sigrral level 2) Noisc 3) Crosstalk.
 E-!@-- r
,.8
'9.

Cotnrnunicatbn & E0240z-May:June 2012)

ffinna link,'is.determined by the optical signal to noise
,,ratio SNR delivered.to thb photodet€ptor.
.!

a
-\
SNR = 14 dB. 1

ln pommercial systems 'i

dB, ' ,'

' 't \'

r\
SNR = 18- 20 *'

these values then determine the: I i

1) Amount'of optical powerthat must be launched into each wavelength
channel.

I
 ./
B,E./B,Tech 5 Examinatio! (8C2402-May-June 201 2)
-:
Interchannel':croistalk arises wlien an interfering sjgnal comes
'

from a neighbouring channel thaloperates.at a differerrt wavelengh.

Fig. shows an example of crosstalks in a demultiplexer.

wavelength'as the desiied signal.

i

This effect is rnore severqthan interchanuel'crosstalk, since tlre,

interference fuils completely within the receiver handwidth. .

Fig. gives tlrg example of an origin of intrachannel crosstalk.

Srgnal
?t"z

.
Il
- -r Crosstalk
___r i__

q\

lu2

--i-
-1,j -
"; -.[-1-
Outputs
Demux
t-l

Fig, : Example of origin of lnterchanngl Crosstalk in a WDM system

Fig. Origin of intrachannel crosstalK in a WDM system

-il'/

Here two independent signals each at a wavelength tr,, enter an

optical switch' This switch routes the signal entering port i to output
port 4, androutes the signai entering poft-2 to outpu, [o* 3. Within the
switch, a spurious fraeiion of the optical porver entering port I gets
coupled to port 3, where it interferes with the signal from-p ortZ.

PowerPenaltiesduetoIntrachannelCrosstalk
If the average received interchannel crosstalk power is a
 i1 l::
.

Op&alCornmur*rttmtt{etworking (E0fl02.May"}une2012), .,-5?

metion E of the awrage receivedlippal power B then in an arnplified
. ' system, where the dominant noise cd$rponent is signat dependent.

T'he intracha,nnel powsr penalty it. i' :

ai
,. Je)
:

If there are N interfering channels in

t'

A WDM, system, each

-t

contributing an average crosstalk power sp, then the fraction e i$, I

,.
given by
ra

{ r- .'

Vti
t

-,\
Let'the received mosstalk power he'a fraction I of the average

Considering an amplified system!

,:
Penalty infer = * 5 log ( I - t)
'i: t
;.t .

where

'rttr**