CHAPTER I

Overview of Optflcal Fiber

Conmmrnlmflaatioms

Ever since ancient times, people had a principal need to communicate with one another. This need
created interests in devising cornrnunication systems for sending messages from one distant place to
another- Optical communicition rnethods were of special interest among the many systems that
people
tried to ori. On, of the eafliest known optical transmission links was a fire-signal method used by the
Greeks in the eighth centgry,BC for sending alarms, ealls for help, or anRouncements of certain events.
Improvements oi these optical transmission systems were not pursued very actively because of technology
limitations at the time. For exampte, the speed of sending information over the communication link was
limited since the transmission rafe depended on how fast the senders could move theirhands, the optical
signal receiver was the error-prone human eye, Iine-of,sight transmission paths were required, and
atmospheric effects such as fog and rain made the transmission path unreliable. Thus it turned out to be
fastelmore efficient, and more dependable to send messages by a courier over the road network.
Subsequently, no significant advances for optical communications appeared until the invention of
the laser in the early 1-9@s and a series of technology developments related to optical fibers around
1970. These events fiaally.allowed practical lightwave communication systetns to start being fielded
worldwide in 1978. These systems operate in the near-infrared region of the electromagnetic spectrum
and use optical fibers as ttgr transmission medium. The goal of this book is to describe the various
technologts, implementationmethodologres, and performance measurement techniques that make optical
fiber corimunic;tion syste.ms possible. The reader can find additional information on the theory of light
propagation in fibers,.the derign of links and networks, and the evolution of optical fibers, photonic
devices, and optical fiber communication systems in a variety of books and conferen"" pro.""dingr''''
This chapter gives an overview of fundamental communications concepts and illustrates how optical
fiber transmissiJn systems operate. First, Sec. 1.1 gives the motivations behind developing optical fiber
transmission systems. Next Sec. t,2defines the different spectral bands which describe various operational
wavelength iegions used in optical communications. Section 1.3 explains fundamental data
communicationioncepts, encoding methods, channel capacity, and the decibel notation for expressing
optical power levels. Section 1.4 gives the basic hierarchy for multiplexing digitized information streams
uied on optical links and Sec. t.5 describes how wavelength division multiplexing can boost the
 transmission capacity of an optical fiber significantly. Next, Sec. introduces the functions and
implementation considerations of the key elements used in optical fiber systems.
An important aspect of realizing a smoothly interacting worldwide lightwave network is to have
well-established international standards for all aspects of components and neiworks. Section 1.7 discusses
the organizations that are involved with this standardization activity and lists the main classes of standards
related to optical communication components, system operations, and installation procedures. Finally,
Sec. 1.8 gives an introduction to modeling and simulation tools that have been developed to aid in the
design of optical fibers, passive and active devices, links, and networks.
Chapters 2 through 10 describe the purpose and performance characteristics of the major elements in an
optical link. These elements include optical fibers, light sources, photodetectors, passive optical devices,
a{rP_lifie_rs,,and active optoelectronic devices used in multiple-wavelength nitwork.. 'it Chapters
9Pt:"t
I I through l4 show how the elements are put together to form links and networks, and "o
explain measurement
methodologies used to evaluate the performance of lighrwave components and links.

t* Motivations for Lightwave Communications

a

Prior to about 1980 most cornr,nunication technologies:involved some type of electrical transmission
mechanism. The era of electr,ical communications star,ted in 1837 with.the invention of the telegraph by
Samuel F. B. Morse. The telegraph system used the Morse code, which represents letters and numbers
by a coded series of dots and dashes. The encoded symbols were conveyed by sending short and long
pulses of electricity over a copper wire at a rate.of tens of pulses per second. More advanced telegrapl
schemes' such as the Baudot system invented in1874, enabled the information speeds to increise to
about 120 bits per second (b/s), but required the use of skilled operators. Shortly thereafter in 1876
Alexander Graham Bell developed a fundamentally different device that could tansmit the entire voice
signal in an analog form and which did not require any expertise to use.24,25
Both the telegraph and the analog voice signals were sent using a baseband transmission mode.
Basebandrcfers to the technology in which a signal is transmitted directly over a channel. For example,
this method is used on standard twisted-pair wire links running from an analog telephone to the nearest
switching interface equipment. The same baseband method is used widely iroptical communications,
that is, the optical output from a light source is turned on and off in response to the variations in voltage
levels of an information-bearing,electical signal.
In the ensuing years an increasingly larger portion of the electromagnetic spectrum was utilized to
develop and deploy progressively more sophisticated and reliable eleitrical communication systems
with larger capacities for conveyinginformation from one place to another. The basic motivations.behind
each new system application were to improve the transmission fidelity so that fewer distortions or errors
occur in the received message, to increase the data rate or capacity of a communication link so that more
information can be sent, or to inciease the transmission distance between in-line repeater or amplification
stations so that messages can be sent farther without the need to restore the signai amplitude;r fidelity
periodically along its path. These activities led to the birttr of a wide variety of communication systems
that are based on using high-capacity long-distance terrestrial and underseacopper-based wire lines and
wireless radio-frequency (RF), microwave, and satellite links.
t
r--
I

i-
Oueruieu of Optical F-lber Commnications

In these developments,tfte basiro ffend.for advancing the link capacity was to use increasingly higher
channel frequencies. 16s leassn,for this trend is that a,time:varying baseband information-bering
signal rnay be transferrod oy:€r &,cofllrnunication channel by superimposing it onto a sinusoidal
electromagnetic wave, which is,known,as the carrier wave or silmply canier. At the destination the
baseband inforrnatio4 signal'rg.removed from the carier wave 'and'processed as desired. Since the
amount of information that can be transmitted is directly related to the frequency range over which the
carrier operat€s,, incleasing.the carrier frequency theoretically increases the available transmission
bandwidth and consequentf, provides a larger information capacity.zfl8 For example, Fig. 1.tr shows

Designation. , Transmission media Applications

1015 Hz

1016 m

I01a Hz

Millimeter 100 GHz

waves
Navigation
Super higft Wavesuide I Satellite-to-satellite
,frequency Mrcrowave Microwave relay 10 GHz
, ($m) - radio Earth-to:satellite
l0 Radar
Ultra high"
fregieacy'r I GHz
. (UHF) r: j, I]I{F TV >l
9pl Mobile. Aeronautica
o
g
o 9
o Veryhigh
6 frequency; Shortwave VHF TVandFM l00MHz S
(VHF) radio rL
Mobile radio
10 m
HiCh Business
frequeh.y Amateur radio 10 MHz
(I$) lntemational
100 " Citizen's band
Medium
frequency AM broad casting I MHz
(MF)
Longwave Aetonautical
Low radid
Aequency Submarine cable 100 kHz
,,,:(I"F), Navigation
l0 Transoceanic radio
'Ve4rlcw.- .,,Wire
frequenc,y, pa!!s 10 kIIz
(vLF)
I
100 Telephone
Audio I Telegraph
+
I kHz

i fE. 1.1 Tle regions oJ the etectronryneticspec'tntm used Jor rodio and optical fiber
I
I
clrlmmwrdlcntions. f{/sed uith permission Jrom A. B. Carlson, Communication
rystems, @ 1986, M&ratu-Hilt B@k CompanA.)
I

1
theelecmtragretie spffil bandsmed fer radio transmission As the diverse r,adio tecbnologies move
from high frequeaey (FIF) to very-hig\frequenc)r (VHF) to ulra high frequency (UFIF) bands with
norniuatr curier frequencies of IO', 10P, and tA' Hz, respectively, increasingly higher infurmation
transmissie* speeds can k en$oyed to provide a higher link eapacity. Thus &e trend in eleetrical
sorrrnunication systefir developrrent$ was to, use prcgressively higher frequencies, which offer
comesponding ilrereases ia badwidth or inforrnation eapaeity-
As Fig 1 .1 also shows, otricat firque,rcies are several orders of magnittrde higher than those used by
electrical comrnunication systeers. Thus the invention of the taser in the early 196Os aroused a curiosity
about the possibility of using the optical region of the electromagnetic spectrum for transmiuing
information. Of particulm interes is the near-infrared spectral band ranging from about 77O to 1675 nm,
since this is a low-Ioss regior* in silica glass fibers.lE-2o The technical breakthrough for optical fiber
eommunications started in 1970 when researchers at Corning demonstrated the feasibility of producing
2e
a glass fiber having an optical power loss that was low enough fbr a praetical transmission link.2o'
As research progressed, it became clear that many complex problerns made it extremely difficult to
extend the carrier coficept for achieving a super broadband optical communication link. Nevertheless,
the unique properties of optical fibers gave them a number of performance advantages compared to
copper wires, so that orptical lirks operating fur a simple on-off keyed baseband mode werc attractive
applications.
The first installed optical fiber tinks which appeared in the late 1970s were used for transmitting
telephony signals at abclt 6 trrfbls over distances of around l0 km. As research and development
progressed, the sophistication and capabilities of these systems increased.rapidly during the 1980s to
create links carrying aggregated data rates beyond terabits per second over distances of hundreds of
kilonreters without the need tc restore signal fidelity along the path length.
Starting in the 1990s there was aburgeoning demand on communication-network assets for bandwidth-
hungry services such as database queries, home shopping, highdefinition interactive video, remote
education, telernedicine ard e.health, high-rriglgdon editing of home vidoos, blogging, and large-scale
high-capacity e-science and Grid computing.3G33 This demand was fueled by the rapid proliferation of
personal computers (PCs) coupled with a phenomenal increase in their storage capacity and processing
capabilities, the widespread availability and continuous expansion of the Internet, and an extensive
choice of remotely accessible programs and information databases. To handle the ever-increasing demand
for high-bandwidth servises froxr ranging from home-based PC users to large businesses and research
organizations, telecomrnunication companircs worldwide greatly enhanced the capaeity of fiber lines by
adding more independent signal'c4q1dng wavelengths on individual fibers and increasing the transmission
speed of inforrnation being carried by each wavelength.

The advantages of optieal fitlers eompared to copper wires include the following:
Long Distance Transmhsian Optical fibers have lower ffansmission losses eompared to copper
wires. Consequently data can be sent over longer distances, thereby reducing the number of
intermediate repeaters needed to boost and restore signals in long spans. This reduction in equipment
and components decreases system cost and complexity.
Large Information Capacity Optical fibers have wider bandwidths than copper wires, so that
more information can be sent over a single physical line. This properry decreases the number of
physical lines needed for sending a given amount of information.
r
I
1-

A:a uAo gJ.Oedcal Flbq furr,lors:unicafinrrs

Srnall Size ad law We@fit The low weight and the srnall dimEnsions sf fibers offer a distinc.t
advantage over heavy, brdky wix'e cables in crowded underground city ducts or in ceiling-mounted
of irnportance in aircraft, satel{ites, and ships where small, lighMeigts
cable trays. This feature also is
cables are advantagmus, ard ia trctical military applications wllere large amounts .of cable nrust
be unreeled and reuieved rapidy.3a
Imm*niq tu Elactt*:dl lac{wnn An especially important feature of an optical fiber relates
to the fact fiat it is a dieleetic material, which means it does not conduct electricity. This makes
optical fibers immurrc to Sre el€cfomagnetic interference effects seen in coppff wires, such as
inductive pickup from other adjacent signal-carrying wires or coupling of electrical noise into the
line from any type of reuby eryipment.
Enhanccd fulcly Wical fibers offer a high degree 'of operational safety, since they ds not have
the problems of ground loops, qparks, and potentially high voltages inherent in coppa lines. Hswever,
precautions with respect to ligh emissions need to be observed to prevent possitle eye damage.
laser
Incrcascd Signal Secarity An optical fiber offers a high degree of data security, since the optical
signal is well-confinod within the fiber and an opaque coating around the fiber absorbs any sigaal
emissions. This feature is in contrast to copper wires where electrical signals potentially could be
tappd off easily. Thus fibers are atractive in applications where information security is important,
such as financial, legal, governmeal, and military systems.

l1"2x Opttcal spectral Baads

All *leommuaicatirm sys$ems use ffi ferr ofelecuomagnetic enerry to transmit signals. Tk spectrron
.fum in Aig. L2. Elecoonugaetb erzrgy,h of
of electrcmftgnetic {EM)
"*l**m, b a combiaation

Fiber optics
to 1675 nm
770
(-375 to 176Tl{z\

IIFIUI$ruISBadiio ffic*owaves
Iafrared ligbt
'1,-7 umts I
Frequancy
,w)
1012

Photoneaugy(eV)

Wavelengtb
I (n)
I l0-2 l0-4 10-6 l0-8
rtg. 1.2 Tte specfram of elrcfiomognetic radintion
 microrryaves,:infrared lighL visible light'
electrical and magnetic fields and.includes power, radio waves,
g*u ,uys. Each discipline takes up. a portion (or band) of the
electromagnetic
ultraviolet light, x rays, *O
is that it can be'viewed as
spectrum. The fundamentll.natureLf an raaiation within thisipectrum
aboyt c = 3 108 m/s in a vacuum' Note
x
electromagneti" *ur", th*t travel at the speed of light, which is
factor r than' the speed c in a
that the speed of light s in a material is smaller U-y tfre refractive-index
vacuum, as chapter 2 descriles. For example , n ='1.45 for
silica glass, so that th9 speed of light in this
material is aboui s = 2 x108 m/s.
'"'i;;;;;ffi;r;ff* ,h; waves in different parts of the spectrum can be mealyred in several
"f
irtioJut"i *uyr. fir"r" art tt " length of one period^of 1h9 waye, the energy conlained
in the wave' or
to.use frequency.to
il;;1il;, i."n*"* of the wave. whereas electrical signal transmission tends to designate the
generally nseswavelengrh
designate the signal operating bands, optical communication
operating region id photon ener_gy or optical poiff when
discussing topics such as signal
spectral
,il"ngtt oi component performance'
"t.rtti-oitical the physical properties of a
As can be seen mm rig. 12, there are three different ways to measure
units are 'related by some simple
wave in various ,"gio;r^ii #'EM tp;;,*"r. These meas,r€ment frequency
to the wavelength times the
equations. First of AL in a vacuum the speed of light c is equal '2'

% so that
c -- hv (l'1)
hertz (Hz).
where the frequency y is measured in cycles per second.o t
energy of a pioton and its frequency (or wavelength) is determined by
The relationship between the
the equation known as Planck's law ' - " :
E= hv 0'2)
constant' The unit J ioules and
where the parameter h = 6.63 x 10-3a J-s = 4.14eV-s is Pl.anck's
mewrs
(measured in units of pm), the energy in
the unit ev stands for electron volts.lnterms of wavelength
electron volts is given bY
r.2406 (1.3)
E(eV)=;-.-----
L\pm)
ultraviolet'"4o1t for1 mm
Figure 1.2 shows, the optical tp".tto* ranges from about 5 nm-11the
visible band. optical fiber
far-infrared radiation.:IJi"i*"", these liirits is the 400-to-700-nm
from nominally 77O to 1675-nm'
communications use the near-infrared. spectal band tangqng
(ITU) has designated six spectral bands for use in
The International Telecommunications,Union
,egior.3s These long-wavelength band
optical fiber communi"u,i*, within the 1260-to-t6i5-nm
of optical fibers and the performance behaviour
designations arose from the attenuation characteristics
respectively' Figure 1'3
of an erbium-doped m"r-"-piin"t @DFA), which Chapters 3 and 10 describe,
shows and Table 1.1 O"n ,"giorr. *tti"t are known by the letters O, E, S, C, L, and U'
"r'tfr" fiber systems. Thus this region is
The 770-to-g10 nm band is used for shorter-wavelength multimode
d".i;;;J;r ,rr" ,ioi-iorrt"ngth or multimode fibel band. Later chapters describe the operational
components, and otherpassive
performance characteristics and af,plications of optical fibers, electro-optic
'optical bands.
devices for use in the short- and long-wavelength
 Oueruiew of Optical Fiber Commwrjcatians

l :U$ed in opeal$ber,comrnitn&.&rs

Designation Spectrum (nm) Origin of Name

Original band O-band 1260 to 136O Original (first) region used for single-mode fiber links
Extended band E-band 1360 to 1460 Link use can extend into this region for fibers with low
water content
Short band S-band 1460 to 1530 Wavelengths are shorter than the C-band but higher
than the E-band
Conventional band C-band 1530 to 1565 Wavelength region used by a conventional EDFA
Long band L-band 1565 to 1625 Gain decreases steadily to 1 at 1625 nm in this longer
wavelength band
UltraJong band U-band 1625 to 1675 Region beyond the response capability of an EDFA

O-Band E-Band S-Band C-nandl t-Band lu-eana

1260 1360 t460 1530 1565 t625 1675

Wavelength (nm)

Fig. l.B Desigrutions oJ spectralbands usedJor opficalJiber @ffmlunico.tiotTs

l

m f'undamental Data Communlcation Concepts

The exchange of information between any two devices across a communication channel involves using
some type of electrical or optical signal which carries this information. The channel could be a wire,
radio, microwave, satellite, atmospheric infrared, or optical fiber link. Each type of channel medium has
unique transmission performance characteristics associated with it, which ideally should match the
properties of the signai. However, regardless of its type themediumdegrades the fidelity of the transmitted
rignut U".uose of an imperfect reproduction of the signal format and because of the unavoidable presence
ofllectrical and opticai noise and interference. These impairments can lead to misinterpretations of the
signal by the electronics at the receivingend.

1.3.1 ElemeTtqr.y Commrurication Link

Figure 1.4 shows a block diagram of an elementary communication link. The user or device where the
-Jr.ug" originates is called i rouru and the final receiving user or device
is called the destination.The
output-of tlie information source serves as the message input to a transmittel' Tl" function of the
traisminer is to couple the message onto a transmission channel in the form of a time-varying signal
that matches the transfer properties of the channel. This process is known as encoding.
As the signal travels through the channel, various imperfect properties of the channel medium and of
various link-components induce impairments into the signal. These include electrical or optical noise
I effects, signal distortions, and signal attenuation. In the presence of these impairments, the function of
the receiveris to extract the weakened and distorted signal from the channel, amplify it, and restore
it as
close as possible to its original encoded form before decodirzg it and passing it on to the message destination.
 AAAT
tttr
Message Encoded Weakeued Restored
the signal input and distorted and decoded
input to
transmitter to the channel received signal message

ng. 1.4 Tfe n:rcrin ompnents bt an elementarg commtutirrrtlon ti,rir-

1.3.2 Analog Slgnals

The format of a signal is an important factor in efficiently and reliably sending the sigaal across a
network. The signals emittgd by informatioR sources can be classified into analog and digital formats.
An analog signal conveys information through a continuous and srnooth varintion in time of a physical
quantity such as optical, electrical, or acoustical intensities and frequencies. For example, voice, music,
electronic hum, and video inforrnation strearis are well-known analog signals.
A fundamental analog signal is the periodic sine wcve, shown in Fig. 1.5. Its thrce main characteristics
are its amplitude, period or frequency, and phase . The amplirude is the size or magnitude of the waveform.

llf Period = lf llf

Ffg. 1.6 llequenq, pdod" ard" amplilude ctnraxtertstics oJ abastc silte wole

Amplitude is designated by the symbol A and is measured in volts, amperes, or watts, depending on the
signal type. Thefrequercy (designated by/) is the number of cycles per second that the wave undergoes
(i.e., the number of times it oscillates per second), which is expressed in units of hertz (Hz). A hertz
refers to a complete cycle of the wave. The pe riod of awave (represented by the symbol 7) is the inverse
of the frequency, that is period = T = llf. The terrr phase (designated by the symbol @) describes the
position of *re waveform'relative to time zero. Phase is measured in degrees or radians, where 180
degrees = zradians.
If the crests and roughs (high and low points) of two waves u,ith identical periods are aligned, they are
said to be in phase.For example, wave I and wave 2 shown in Fig. 1.6 are in phase. l.et wave t have an

Exam?le 1,1 A sine wave has a frequency/:= 100 A sine wave is offset by 1/4 of a cycle with respect to
kHz. Its pedod is I=/lOJ s = 0.01 ms. time zero. since one cycle if 360 degrees, the phase shit is
A sine wave has period I = I ns. Its frequency is / = Q = 0.25 x 36O degrees = 90 degrees = nlZ radran
qs)
l/( t0 = I GHz.
amplitude Ar and let wave 2 have an amplitude
Ar. If these two waves are added together, the
amplinrde A of the resulting wave will be the sum
A = At+ Ar. This effect is known as constructive
interference.
Figure 1.7 illustrates four phase shifis of a
wave relative to time zero. When two waves
differ slightly in their relative positions, they are
said to be out of phase. As an illustration, the
wave shown in Fig. 1.7(c) is 180' (zrradians)
out of phase with the wave shown in Fig. 1.7(a).
If the two waves in Fig. 1.7(a) and 1.7(c) have
identical periods, the resulting wave amplitude
will be the sum A = Ar + Ar.lf the two waves
also have the same amplitrr,rles, ithen they undergo
destructive interference so thatA = 0, that is they
will cancel each other out. These interference
concepts are important when considering the A1+ A2
operation of devices such as laser diodes, thin-
film filters, and optical cbuPlers. fig. 1.6 Two in-pha,se u)aues usith tlw same
perid will add" crn:.s,huctitselg

ll4 cycle o) 90'

ll2 cycle

Addition of two waves that are

7r radians out ofPhase Yields
zero final amplitude

fig. 1.2 lltustationoJJour phase shifis oJausaue reWitse to time zcro. Ttuo identicollDaues
that are 78O" ortt oJ ptwse will interJere destrucfwelg.
Example 1.2 If the spectrumof a signal ranges from la1 = 100 llllzr then tk bandrryidth' B =f*** fw,,=90 kHz'
its lowest frequencyJ"* = 10 kllz to its highest frequency
,.,

(o.rsimpli
Two further co]nmon characteristics in communications are the frequency spe:trum
spectrurn) and the b;;iil ;f a signal. The spectrun+ of a signal'i: r-n:.*q*:f fteQuencies that it
sine waves of different
conrains. That is; the .p"it u; of u silnal is the combination of At tfre individual
f,rdquencies which mak!-64;, .igi;r. T\e _bandwidrh (designated by B)
refers'to.ths width of 'this
(kI'Iz), megatrertz (Iv{I{zi' or
,p""*-. ThelqDdrvidth .otqsllgay. is specified in units,such as kilohertz
gigahertz (GHz).

1.3.3 DiEitat Sfgnals

A digital signal is an,ordered sequenceof discrete symbols' selected from a finite set of'elements, for
@, #;'or'vo: A eoaffdn disital
;;ffi;; ffih6o""f}j.-niil;#;;;;b;, "rd other symbols socfi aspulse
twd types of shape'as
signal configuration is ttw. binary waveprm, wnich consists
:l: --nown
il tt" of
;f; in Fii. 1.8, Th€ information contained in a digital slsnat swT by
knyWrzey,,?r
Partic.ulal segrlettce,
h"t."'
the presence (abinarry one,;orsirnply either one or lfand absxiea(a :ly-pll"i
loqic l) atfi a lagic Tero:(ot
or 0) of these pulses or D;{s, These iiis oten are referred to as aIoi*
one
,(ot
logi.c0),respictivelY.r.. ,., l

-*1 F-
Tt =llR:bitinterval
(a)

-, l*-
Bit duration To:bit interval

(b)

Ftg. 1.8 Exanqtbs oJ binary uaueJorms

tttso tleir amplitude: wn4, and bit
'r'r,yiy
dwqtiotl" (o) fi\e bitfrlls tne eittre bit pe:riod.Jor ttrc t b-it ont4;
(b) a 1 bttJills the
slot
fvst W qrtd.a,O bitfitls the seond ha{ of a
bit frin7

The time slot fi,,irr urhich abit occurs is called either the bit interval,bit
period, or bit time' The bit
a'rate of R bits'per iecona'6ts\'ryhere
intervals are,spdcsd reguhr{y,ald occur every 1/R secoirds,:or'at
R is the bit rate o, tn" iotiiote.Tltedata rate commonly ii specffied in units such asrkilobits per second
 tai:!W
Oueruiew oJOpticalF-tber Commtutications ,,:H ,

(kb/s), megabits per second (Mb/s), or gigabits per second (Gb/s). As an example, a datarate of 2 x 10e
bls = 2 Gb/s. A bit can fill the entire bit interval or part of it, as shown in Fig. 1.8(a) and Fig. 1.8(b),
respectively. A block of eight bits often is used to represent an encoded symbol or word and is called an
octet ot a byte.

1.3.4 Dtgitizatton of Anatog Signals

To send a signal in an analog format, the transmission channel typically cannot achieve perf'ect
reproduction of the signal at the destination, so there always will be some degree of distortion.
Furthermore, different,types of analog signals may require different channel responses. This puts a
strain on the design of amultipurpose analog link. In addition, noises that get added to the analog signal
as it passes through multiple repeaters cannot be eliminated, which further degrades the signal.
In contrast, digital signals can undergo a gteat amount of distortion and still allow the information to
be exfiacted with a higf, degree of fidJity. to avoid the shortcomings of analog signals and to create
networks that can multiplex and switch any type of information, most informatipn is sent in a digital
format. To achieve this network flexibility, most analog signals need to be converted to a digital format.
However, as Chapter 9 describes, there are a number of situations in which it is advantageous to send high-
speed analog signals in their native form over relatively short distances. One example is the transmission
of microwave signals from a satellite dish to a processing station located less than a, kilometer'away.
An analog signal can be transformed into a digital signal through a process of periodic sampling and
the assignment of quantized values to represent the intensity of the signal at regular intervals of time. To
convert an analog signal to a digital form, one starts by taking instantaneous measures of the height of
the signal wave at regular intervals, which is called sarzpling the signal. The simplest, but not necessarily
the bJst, way to convert these analog samples to a digital format is to divide the amplitude excursion of
the analog iignal into i/ equally spaced levels and assign a discrete binary word to each level. Each
analog sample is then assigned one'of these level values. This process is known as quantizatian. Since
the signal varies continuously in time, this process generates a sequence of real numbers.

Examole 1,3 Figure 1.9 shows an example of chosen as the one to be transntitted, according to the 3-bit
digitization. Here the allowed voltage-amplitude excursion binary code listed next to the quantized levels shown in
is divided into eight equally spaced levels ranging from Fig. 1.9. At the receiver this digital signal is then
zero to V volts. ln this figure, samples are taken every demodulated. That is, the quantized levels are reassembled
microseiond and the'nearest discrete quantization level is into a continuously varying analog waveform.
Binary
code
number
111
6 6 110
101
=
9q
Ec 100
011
010
2
001
000
r0 t2 l0 12
(a) (b)
Ftg. 1.9 Concept Jor digitizatian of anahag signals. (a) Originat signal uarying betuseen O
and. V wlts: h) tumpled and qtntttized d$ital uersion
 Intuitively, one can see from Fig. 1.9 that if the digitization samples are taken frequently enough
can be recovered
relative to the rate at which the sign-al varies, then to a good approximation the signal
points. The resemblance of the reproduced
from the samples by drawing a stralght line between the sample
signal to tfre iriginat .ijnuia"p"nd-s on the fineness of the quantizing process and on the g{ect of noise
and distortion added irito the Lansmission system. According to the Ny4aist the9r9;1t, if the sampling
device can faithfully ieconstnrct the
rate is at least two times the highest frequency, then the receiving
if a sigial is limited to a bandwidth of Bhertz,then signal can be reproduced
;;;;;d;;iirtTh;, the
by
without distortion if it is sampied at a rate of 28 times per second. These data samples are represented
a binary code. As noted in iig. t.l, eight quantized levels having upper
bound? Vr: Vr: "" V can be
give finer sampling levels. That is, if n
described by 3 binary Agrrr (ZY = S1. Vt-ore &gits can be used to
binary digits represent each saraple, then one can have 2" qtanization levels'

1.3.5 Channel CaPaelty

This is the
m ,fr(, ,r"fysis of any com,rnunication network, an important factor is channel capacify'
to the user destination.
maximum rate at which data can be sent across a channel-from the message source
theorem states that
A fundamental and important theorem for this is the Sharuwn capacity formula.Tlis
capacity
if a channel has a bandwidth B (measured in hertz) then the maximrrm information-transmission
C of that channel is given trt bitS per secondby the relationship6-28'36
iBtogz(l + S/If) (1.4)
C
Here log2 r"pr"t"Utq the base-2 logarithm, and S and N are the average
signalpower and noise powel'
point that t1gn4
,"rp"qtiu"ff fypicaUy'these powex are measured at the ryc.elv* since it is at this i it
Nore: For simplicity of calculation, the following relationship
extracted f;om fie channel'ani proeessed.
-QL
may be useful to find log2 "r:
x)t0'3. (1'5)
logz v = (lo916 x)t(log*2) = (lo916
power in a signal to the
The parameter S/N is the signal-to*oise ratio (SNR), which is the ratio of the
expressed in decibels:
power contained in the noise a:t a particular measurement point. This ratio is often
(1.6)
sNRdB = ro los
H#ffi = ro ros
f,
achieved. In practice
The Shannon formula indicates the theoretical maximum capacity that can be
into account thermal noise and does not
this capacity cannot be reaehe{ since the formula only takes
delay distortion. Furthellnore' intuitively
consider factors such as impulse noise, attenuation distortion, or
by raising the signal strength. However, raising
it might seem that the capacity can be increased merely
leads to highel noise powers' Also
the slgnal level also inciJasei'nonlinear effects fn the iysiem,_wt-rich
ratio S/N, since the wider the bandwidth the more
,"," ,irri increasing the bandwidtb B decreases the
noise is introduced into the system.

signal-to-noise ratio is 20 dB. Then the bandwidth is

Examole 1.4 Supposewe'haveanoisychannelwith
a l-MHz bandwidth in which the signal-to-noise ratio is 1' B = (4MIlz)- (3 MHz) = 1 MHz
From Eq. (1.4), the maximum.capacity for this thanpel is and, from Bq. 11.6).
C =-Blogz(1 + S/N) - 106log, (1 + 1) = 106log, (2) 571g_1920/ro=100
= 106 11.0; = 1 Mb/s Then
C - (1 + 100) = [106 logro (101)]/0'3
106 log2
Example 1.5 Let us,find the capacity of a channel
that operates between 3 MIIz and 4 MHz and in which the = 106 (2.0)/0.3 = 6.7 Mb/s
 Anruieu: oJ Opfiml F'Ibr bmmnications

1.3.6 Declbel Unlte

mechanisms in atransmissionmedium'
Reduction or attenuation of signal strengtharises from various loss
For example, electric power Is lost thriugh heat generation as an electric signal flows along a wire, and
processes in a glass fiber or in an atmospheric
optical power is attenuaied ttnough scatteing anO iUsorption
periodically along a channel path to
channel. To compensate for the[ energy loises, amplifiers areused
boost the signal level, as shourn in Fig. 1'10'
Original Amplified
signal signal
Attenuated
signal

]-l
Amplifier

ru. 1.1O Paiodicaltg placedamptifiers compensalefor energg losses alarE alink

a link o1-a device is to reference
A standard and convenient rnodrod for measuring attenuation through
optical fiber' the signal strengh
the output signal level to the input level. For guidfo media such
as
P it in terms of a logarithmic
normally decays exponentially. Thus, for con-venience one can designate
power ratio miasured in decibels (dB). The dB unit is defined by
I

t (1.7)
I
Power ratio in dB = 10 r"f
i
{r.
i
where Pr and P, are the elecrical or optical power P1
levels of a signal at points I and 2 in Fig' 1'11, Pz= 0.5Pt
and log is the base- 10 logarithm. The logarithmic
naturs of the decibel allows a large ratio to be
expressed in a fairly simple manner. Power levels
ditrering by many orders of magnitude can be
[:l r_l
I compared easily when they are in decibel prm'
Anoiher attractive feature of,&edecibel is that to
measure the changes in ttre sfength of a signal,
one merely adds or subtracts th decibel numbers Point I Point 2
between two different Points.
Ftg. 1.11 F,:arnple oJpttlse attenuntionino'
tirlk P l and.P2are ttw Pwerleueb
l_. oJ a stgnal at Polnts 1 and 2

i'
I
 ..: :. j -'.:::";--::: : t;. .:.

ExamPle 1.6 airo-" thar after traveling a certain Thus, -3 dB (or a 3-dB attenuation or loss) means that
distance in sorire translnissibn rntdium,..tlrepower'of asignal the signal has,lost half its power- If,.an anplifier'is inserte.d
is reduced to half; that is, P2 +Q5 f1riaFig. t;l1. At this into the.ltrk at:rhis point to boost the signal back to its
poinq using,Eq; (1:7) the *tieu@bnorr less of,powe{ is origtgal,lgveJ tren,that amplifrer has 4-3-dB gain. H the
amplifier has a 6-dB gain then il boosts the signal.power
O't-4
r0log
"n9 =t0log" Pl
=I0log0.5=10{-0.3)=-3dB level to twicg tbe origrnat value.

Table 1.2 shows some sarnple values of power loss

given in decibels,and drrc percent of power remaiuing
after this loss. These types of numbers are innportant
when consideringfacbrs sueh as the effects of tapping
off a small part of an optical signal for monitoring Power loss Percent of
purposes, for examining the power loss through some (in dB) power left
optical element, or when calculating the signal
attenuation in a specific length of optical'fiber. 0.1 98
o.5 89
1 79
2 63
J 50
6 25
10 10
20 1

Example 1.7 Consider the transmission path from dB levei at point 4 - (loss in line 1) + (ampliher gain)

point I to point 4 shown in Fig. 1.12. Here the signal is + (loss in line 2)
attenuated by 9 dB between points 1 and 2. After getting a , r (-9dB)+(14dB)+(-3 dB)=+2dB
14-dB boost from an amplifier at point 3, it is again Thus the signal has a2-dB (afactor,of 100'2 = 1.58)
attenuated by 3 dB between points 3 and 4. Relative to I to point 4.
gain in power in going from point
point l. the signal level ih dB ar point 4 is

Transmission line 2

Psint I Point 2 Point 3

,:: i. -;-':': .. - : : :,,

E!g, i,.1?.. Nifiitb oJ:rgitsl atterunfan atd. ompt!fication in d transmissfon Wfh

 aueruiew of Optigal Fiber Commwtiaqtions

Since the decibel is used !9 rgf..-g{ 10 Iatlos or relative

units, it gives no indiCationofifhe absolute power level.
i
However, a derived unit can be used for this. Such a
unit that is particularly common in optical fiber
Power dBm equivalent
communications is the.dBrn (simply pronounced dee
bee em). This unit expresses the power level P as a 200 mW 23
logarithmic ratio of P referred to 1 mW. In this case, 100 mW "20
the power in dBm is an absolute value defined by 10 mW 10

r li" lmW 0
i.-
'n
Power level (in dBm) = 161o* tY) (1.8) 100pW -10
10pW -20
An important'ru1e-+f.,thumb relationship to lpw -30
remember for optical fiber comrnunications is 0 dBm - 100 nW -40
= I mW. Therefore, positive values of dBm are greater 10 nW -50
than 1 mW and negative values are less than 1 mW. lnW -60
100 pW _:70
Table 1.3 lists some examples of optical power levels
I 10 pW -80
and their dBm equivalents.
lpW -90

m Network Informatlon Rates

To handle the continuously rising demand forhigh-bandwidth services fro:nusers ranging from individuais
to large businesses and.research organizations, le,Iecommulication companies worldwide are
implementing increasingly sophisticated digital multiplexing techniques that allow a larger number of
indipendent information sfreams to shale the same physical transmission channel. This section describes
some corlmon digital signal multiplexing techniques.

1.4.1 Telecom
fuUt" t.+ gives examples of infofmation rates for some rypical telecom services. To send these services
t' from one user to another, network providers combine the signals from many different users and send the
1.!
aggregate signal over a single transmission line. This scheme is known as time-diuislstx-ruuh;p-lexing
dbt tl wherein N independertt infomiation stteams, each runriing at a data rate 9f R b/s, are interledved
etectricatty into:a si'ngie infor-mrition stream operating at a highlr rate of l/x R bls. To get aidetailed
perspective of this methodology, let us look at the multiplexing schemes used in telecommunications'
'"n'*fy uppfications of nU"r"opti" transmission linhs were mainly for large caplcity telephone lines.
fnese aigitA'tnks consisted of time-division-multiplexed 64-kb/s voice channels. The multiplexing was
Oevel@6a in *re-19@!' is-based'on what is known as the plesiochrorwus digital hierarchy (PDH).
nigor"i.
-The
t: 'shi s tfri;iAgita i*r*arission hierarchy used in the-North fqrerican !el?,Pne network. _
^
funtlamentat'Uulfriihg.Uldak is a'I.544-MbA triinsmission i'ate tnown as aDS'|. ratp, where DS
stands for digital sys;tem.I1 is formed by time-division-multiplexing ffienff-four Voice chanflels, each
digitized at uZ+-UAirrate (rvhich isr"ferr"d to asDS0). Framingbirs, which indicate where an irrformation
uoit rtrut, and ends, are'added':alohg'with these voice channels to yield the 1.544-Mb/s bit stream.
Framing and other conffol bits that.may get added to an information unit in a digital stream are called
d' overheid.bits. At any multiplexlng ievifa signal at the designated input rate is c-ombined with other
input signals at the same rate.
 Type of service Data rate

Videci on demand/interactive TV 1.5 to 6 Mb/s

Video games I to 2 Mbls
Remote education 1.5 to 3 Mb/s
Electronic shopping 1.5 to 6 Mb/s
Data ffansfer or telecomrhuting 1 to 3 Mb/s
Video conferencing 0.384 to 2 Mb/s
Voice (single telephone channel) 33.6to 56 kb/s

Six I
44.736:lvlbls
lnputs 274.l76lvIbls
Seven I
T3
6.312-MbA {7
t;-- multiplexer
inputs 44.736lvIbls
T2
multiplexer
6.312 Mb/s

TI
multiplexer
1.544 }vlb/s
64-kbls
inputs

fE. 1.13 Wifat fa4gmissrior hierarchg used. in the North American tebphone netutork

DSr versus Tr In describing telephone network data rates, one also sees terms_such as.Tl.,_I3, and
so on. Often ttre .terms Ti andDsle (e.g.,,71 and DS1 or 73 and DS3) are used interchangeably. However
there is a subtle difference in their meaning. Designations such as DSl, DS2, and DS3 refer to a service
type, fot example,'a user *ho wants to send information at a 1.544 Mb/s rate would subscribe to a DSl
seruice.,q,UUreviationsSuchas Tl,n,andT3refertothedataratewhichthetransmission-linetechnology
;;"r 6'J;iir*tn"t service overta physical link. For example, the DSI service is Eansported over a
physical wire oi optical fiber using elecrical or optical pulses sent at aTl = 1.544 Mb/s rate.
The TDM scherne is not restricted to multiplexing voice signals. For example, at the DSl level, any
64-kb/s digital signal of the appropriate format could be transmitted T on: of Og Z!_+put.channels
shown in Fig. 1.13. A_s nqted there and in Table 1.5, the main multiplexed rates for Norttr A.merican
applications-are designated as DSl (1.544 Mb/s), DSz (6.312 Mb/s), and DS3 (M.736 Mb/s). Europeln
,nd Japan"r" networks define similar hierarchies using different bit-rate levels as Table 1.5 shows. In
Euro@ the multiplexing hierarchy is labeled El, E2, E3, and so on.
 Ouentiew of Optical Fiber Commrtrdcations

TebIs,I,,$- rp&tfq*r.nul& g,ler .{s- usedinNorth Arrcrica"'fat'w$q qd1.&ffryl

Digital multiplexing Number of 64-kb/s Bit rate (Mbh)

level channels
North America Europe lapan

0.064 0.064 0.064

DS0 1

DS1 24 r.544 r.544

30 2.O48
48 3.152 3.t52

D52 96 6.312 6.3t2

t20 8.448

D53 480 34.368 32.064

672 44.736
r344 91.053
144A 97.728

DS4 1920 139.264

4032 274.176
5760 397.200

r.4.2 SONET/SDH
providers established
With the advent of high-capacity fiber optic transmission lines in the 1980s, service
optical nerwort (SONET) in North America and s,vnchronous
a standard signal format ciled iynchroious
of These standards deflne a synchronous frame
digital hierirchy (SDH) in other parts the world.37'38
traffic over optical fiber trunk lines. The basic building block
,frr.t r" for sending multiplexed digital
and first level of the SONET signal hierarchy is called the Synchronous Transport Signal - lwel l (STS-
1), which has a bit rate of St.g4 lvtUls. Higher-rate SONET signals are obtained
by byte-interleaving N
scrambled and converted to an Optical Carrier - l,evel N (OC-N)
of these STS-1 frames, which then are
.rgrA. Thus the OC-N signal will have a line rate exactly N times that of an OC-l signal. For SDH
Transport Module - Level I
,f.t"*. the fundamental6uilding block is the 155.52-Mbls Synchronous
generated by synchronously multiplexing Ndifferent
fiffrA-f l. Again, higher-rate information streams are
STM-1 sign-als to form the STM-N signal. Table 1.6 shows commonly used SDH and SONET signal
Ievels, the line rate, and the popular numerical name for that rate.

teble 1.6 Cofia7:Ert SDH and.SOI\rET line rates and. their popdar rwneriml rtalT"E

SONET level Electrical level SDH level Line fate (Mb/s) Popular rote non'te

oc-l STS-1 5r.84

oc-3 STS.3 STM-1 155.52 155 Mb/s
srs-12 STM-4 622.08 622 N{bls
oc-12
OC-+8 STS.48 STM-16 2488.32 2.5 Gb/s
STM-64 9953.28 10 Gb/s
oc-192 STS-192
oc-768 STS-768 STM-256 398 1 3. l2 40 Gb/s
m wDM curcepts
The use of wavelength division multiplexing
(WDM) offers a further boost in fiber
transmission capacity. The basis of WDM is
to use multiple sources operating at slightly
different wavelengths to transmit several
indepehdent information streams simultan-
eously over the same fiber. Figure 1.14 shows
the basic WDM concept. Here N independent
optically formatted information streams, each Individual Optical
transmitted at a different wavelength, are fiber lines multiplexer
combined by means of an optical multiplexer fig. f .f4 Basic corrcept oJtoauelengthdiuision
and sent over the sarne fiber. Note that each of ntttltiplexing
these streams could be at a different data rate.
Each information stream maintains its individual data rate after being multiplexed with the other traffic
streams, and still operates at its unique wavelength. Conceptually. the WDM scheme is the same as frequency
division multiplexing (FDM) used in microwave radio and satellite systems.
Although researchers started looking at SIDM techniques in the I970s, during the ensuing years it
generally turned out to be easier to transmit only a single wavelength on a fiber using high-speed electronic
and optical devices, than t6'invoke the greater system complexity called for in WDM. However, a
dramatic surge in WDM popularity started in the early 1990s owing to several.factors. These include
new fiber types that provide better performance of multiple-wavelength operation at 1550 nm, advances
in producing WDM devices that can sepaJate closely spaced wavelengths, and the development of
optical amplifiers that can boost C-band optical signal levels completely in the optical domain.

tr.o I Key Elements of Optical Fiber Systems

Similar to electrical communication systems, the basic function of an optical fiber link is to transport a
signal from communication equipment (e.g. a computer, telephone, or video device) at one location to
corresponding equipment at ariotheJ location with a high degree of reliability and accuracy. Figure 1.15
shows the main constituen{s of an optical fiber communications link. The key sections are a transmitter
consisting of a-light'sourie'and'its associated drive circuitry, a cable offering mechanical and
environmental protection to the opticdl fibers contained inside, and a receiver consisting of a photodetector
plus amplification and signal-restoring circuitry. Additional components include optical amplifiers,
connectors, splices, couplers, regenerators (for restoring the signal-shape characteristics), and other
passive components and active photonic devices.

1.6.1 Ovenriew of Element Applications

The cabled fiber is one of the most important elements in an optical fiber link. In addition to protecting
the glass fibers during installation and seryice, the cable may contain copper wires for powering optical
amplifiers or signal regenerators, which are needed periodically in long-distance links for amplifying
and reshaping the signal. A variety of fiber types with different performance characteristics exist for a
wide range of applications. To protect the glass fibers during installation and service, there are many
different cable configurations depending on whether the cable is to be installed inside a building,
 Oueruieu of Opfical Fiber Communications

Information Optical connecters Information

sources
recipients
Video
Cabled optical fibers
camera
/ \

I
I
rE. 1.r5 Main onstihrcnts
\
Passive or active oPtical devices
(optical filters, couplers, switches)

oJ an opticalfrher coffarulnicdjfions ltnk

underground in ddcts or through direct-burial methods, outside on poles' or under water.

Very low-loss
networks for joining cables
opticai connectors and splices-are needed in all categories of optical fiber
and for attaching one fiber to another.
As a result of installation and/or manufacturing limitations, individual cable lengths will range
from
reel size and cable weight
several hundred meters to several kilometers. Practical considerations such as
used when the
determine the actual length of a single cable section. The shorter segments tend to be
or underwater applications'
cables are pulled through-rlucts. Longir lengths are used in aerial, direct-burial,
hundreds of kilometers long. These cables are assembled in on-
Transoceanic cable length. ur" -*y individual cable sections
shore factories and then-loaded into special cable-laying ships. Splicing together
forms continuous ftansmission lines for these long-distance links.
fiber. The
Once the cable is installed, a transrnitter can be used to launch a light signal into the
and
transmitter consists of a light source that is dimensionally compatible with the fiber core associated
(LEDs) and laser diodes
electronic conrrol and moJuhtion circuitry. Semiconductor light-emitting diodes
are suitable for this purpose. For these devices the light output can be modulated rapidly
by simply
varying the input .u.."nt at the desired transmission rate, thereby producing an optical signal. The
iJnput signals to the transmitter circuitry for driving the optical source can be either of an analog
"tect
or digital io*.-fn" functions of the associated transmitter electronics are to set and stabilize
the source
op"ruting point and output power level. For high-rate systems (usually greater than about 2.5 Gb/s),
an external
Oirect moautation.of the source can lead to unacc-ptable optical signal distortion. In this case,
modulator is used to varJ the amplitude of a continuous light output from a laser diode source.
In the
(126O to
770+o.910-nm region thi light sources are generally alloys of GaAlAs. At longer wavelengths
1675 nm) an InGaAsP alloy is,the principat optical source material.
After an optical signal is launched into a fiber, it will become progressively attenulted and distorted
glass
with increasing distance because of scattering, absorption, and dispersion mechanisms.in.the
which
.;i"ri"f . At d;e destination of an optical fiber transmission line, there is a-receiving device
that detects
interprets the information contained in the optical signal. Inside the_ receiver is a photodiode
the weakened and distorted optical signal emerging from the end of an optical fiber and converts it to an
electrical signal (referred ti as a.piotocurreit). The receiver also contains electronic amplification
devices andiircuitry to restore signat fidelity. Silicon photodiodes are used in the 770+o-910-nm region'
The primary mat€rial in the 1260-to-1675-nm region is an InGaAs alloy.
ftr" a"rigp of an optical receiver is inherently more complex than that of the transmitter, since it has
to interpret the content ofthe weakened and degraded signal received by the photodetector. The principal
 power n::es:effy at the desired data rate to attain
figure of merit for a receiver is the minimum optical
either a given error probauility for digital ,yrt"*, or
a specified signal-to-noise ratio for an analog
p"tfo**19 level depends on the photodetector
system. The ability of a receiver to achieve a certain
of the successive amplification stages in
type, the effects of noir"l,
tn" system, and the characieristics
the receiver.
devices that assist in controlling and
lncluded in any optical fiber link are various passive optical
that require no electronic'control for
g#irgA; figti.igriuf*. Passive devices *" optl.ut components only a nalrow spectrum of desired light'
their operation. Among these are opti.cal filteis that select
into a number of different branches' optical
optical splitters that diiide the power in an optical signal
onto the same fiber (or that
multiplexers that combine signals from two * -o." iistinct
wavelengths
separate the waretengtt s at rfrZ t"""itt g end)
in multiple-wavelength optical fiber networks' and couplers
monitoring purposes'
;;!6r" tap off u ."ri=ui, p"r"entage of-iight, usually for performance
contain a wide range of active optical
In addition, *oO"* Iopf*i.riJu,"A oftical fiber networks
operation' These include light signal modulators'
components, which require an electronic control for their
elements for adding and dropping
tunable (wavelength-selectable) optical titiers, reconfigurable switches'
nodes, i*iuUt" optical attenuators,lmd ontical
wavelengths at intermediate
a fiber' ii becomes greatly weakened due
After an optical signat tras traveled a certairi distance along
up an optical link' engineers formulate a power
to power loss along ,n" nU"t Therefore, when setting The
budget and add umprffierc or,repearers *tr* ,tr". p*tl. for: ei1e1!s the available power margin'
repeate( also will
periodically placed #J), giu: *" optioal signal a power boost' whereas a
"diifi";;
attempt to restore the Jlgrratto its original *-t uf.l
ni* t]o fggo, o$1 rePeatert.**:111111e for signal

f* un ln""o*irri opi"uf-rignul, ui"p"uto p"tfottsphoton'to-electron conversion' elecffical

amplification.
conversion' This process can be
amplification, retiming, pulse shaping, qd then elecfron-to-pfioto' expended a gteat deal of
systems. researchers
fairly complex for high-speed multiwavelength ]h}s,
light power level completely in the optical
effort to develop all-optical,arnplifiers, wniJfr boost the
domain. Optical ampthcation mechanisms for WDM
links include the use of devices based on rare-
lengths JnU* and distribute6 amplification by means
of a stimulated Raman scattering
earth-doped
effect.
system require measurement
The installation and operation of an optical fiber communication
of the constituent components
rcchniques for verifying ,fi;,h" specified performance
characteristics
parameters, system engineers are interested in knowing
are satisfied. In addition to measuring optical fiber such as
the characteristics of passive splitters, r""to,*, and coupleis, and electro-optic components'
"on
sources, photodetectoi., *O oiti"A amplifiers.
Furthermoie' when a link is being installed and tested'
operational parameters ihat should b1 m3asured
include bit error rate,' timing jitter' and signal-to-noise
maintenance
ratio as indicated Uy tt".V.put During actual operation, measurements are needed for
"-.
and monitoring functionsio determine factors such
as fault locations in fibers and the status of remotely

located oPtical arnPlifiers'

1-6.2 trIindows and SPectral Bands

systerns and the qharacteristics of the four key
Figure 1.16 shows the opelating rTge of optical fiber Here the
componenrs of a link ,fia tU"et tigtrt ,our."., photodetectors, and optical amplifiers'
"pii"""f main traditional operating wavelength bands of
dashed vertical lines indicate the centers oi th" thr"e One of the
region, the O-band' and the C-band'
optical tiber systemr,-*hi"h are the short-wavelength shown at the
as a function of wavelength' as
principal characteristic* oiun optical fiber is its atte"nuation
topinFig.l.16.EarlyapplicationsinthelatelgT0smadeexclusiveuseoftheT7o-to.gl0-nmwavelength
 Oueruiew oJ Optical Ftber Commtutications

band where there was alow-loss window and GaAlAs

optical sources and silicon photodetectors operating
window' since
at these wavelengths were available. Originally this region was referred to as the first
absorption by water mo-11-11es' As a result of
around 1000 nm there was a large attenuation ,pit" ao"io
attenuation curve around 850 nm'
this spike, early fibers exhibitei a local minimum in the
metallic impurities in the fiber material' in the
By reducing the .oo""routi"n of hydroxyl ions and
1980s manufacrurers could fabricate optical fibers with
very'low losJes in the 1260+o-1675-nm region'

Optical fibers
1310 nm

)0 900

Optical sources
r7m77/7777V7777
GaAlAsI lnGaAsP
I
I

Optical amPlifiers

m RamanamPlification r7v7v777777V7/77.?V7Vm

PDFA TDFA EDFA

Optical,fiber amplifiers W w4'v7xvvTVvz

1.0
-

i o.s
q
o
&

700 900 1100 1300 1500 1700

. Wavelength (nm)

oJtteJour keg optielfiber tink componerts

rE. 1.16 Ctwracteristlcs ]d oprathg ranges

This spectral band is called the lo ng-wavelength region' il'" glass still contained some water
.SiT:
1400 nm. This spike defined two low-loss
molecules, a third-ord#il"fi;-.'rtk" remalned
aiound
being tn"i""i"a*iiiow centercd at 1310 nm and the thirdwindow centered at 1550 nm'
windows, these the low-loss
These two windows r"* ;caleil the o-band and C-band, respectively. The desire to use
InGaAsP-based light sources and InGaAs
long-wavelength regions prompted the development of
doping optical fibers with rare-earth
photoderecror. tuut .* opJr.* ii r:ro and 1550 nm. In addition.
elernents such as Pr, Th,:and Er creates optical fiber amplifiers (PDFA; TDFA, and EDFA devices).
These devices. and the use,of,Raman amplification gave a further capacity boost to long-wavelength
WDM systems.
Special material-purification processes can eliminate almost all water molecules from the glass fiber
material, thereby dramatically reducing the water-attenuation peak around 140O nm. This process opens
the E-band (1360+o-1460 nm) tr.ansmission region to provide around,100,nm:more spectral bandwidth
in these specially fabricated fibers than in conventional single-mode fibers.
Systems operating at 1550 nm provide the lowest attenuation, but the signal dispersion in a standard
silica fiber is larger at 1550 qm than at 1310 nm. Fiber manufacturers overcame this limitation frst by
creating the dispersion-shifted fibers for single-wavelength operation and then by devising non-zero
dispersion-shifted fiber (NZDSF) for use with WDM implementations. The latter fiber type has led to
the widespread use of multiple-wavelength S-band and C-band systems for high-capacity, long-span
terrestrial and undersea hansmission links. These links routinely carry traffic at 10 Gb/s (OC-192ISTM.
64) over nominally 90-km distances between amplifiers or repeaters. By 2005 links operating at 4O Gb/s
were being installed and ?ield trials of I60-Gb/s long-distance transmission systems were tested
successfuliy.3e-a2

, tz
.,.rcF
v Standards for Opticat Fiber Communications
To allow components and equipment
-been
from different vendors to interface with one another, numerous
intemational siandards have developed.a3-as The three basic classes for fiber optics are primary
standards, component testing standards, and system standards.

fuimary stolndolrds refer to measuring and charaqterizing fundamental physical parameters such
as attenuation, bandwidth, operatioual characteristics of fibers, and optical power levels and spectral
widths. In the USA the main organization'involved in primary standards is the National Institute of
Standards and Technology (MST). This organization carries out fiber optic and laser standardization
work, and it sponsors an annual conference on optical fiber measurements. Other national organizations
include the National Physical Laboratory NPL) in the United Kingdom and the Physikalisch-Technische
Bundesanstalt (PTB) in Germany.

Component t*ting standardsdefine tests tbr fiber-optic component performance and they
establish equipment-calibration procedures. Several different organizations are involved in formulating
testing standards, sorne,v€ry,active ones being the Telecommunications Industry Association (TIA) in
association with the Electrsn_ics Industries Alliance (EIA), the Telecommunication Sector of the
International Telecomrnunication Union (ITU-T), and the International Electrotechnical Commission
(IEC). The TIA has a list of over 120 fiberoptic test'standards and specifications under the general
designation TIA/EIA-455 XX-YY. where XX refers to a specific measurement technique and YY refers
to the publication year.'These standards are also called Fiber Optic kst Procedures (FOTP), so that
TIA/EIA-455-XX becomes FOTP-XX. These include awide variety of recommended lnethodsfortesting
the response of fibers, cables, passive devices, and electro-optic components to environmental factors
and operational conditions. For example, TIA/EIA-455-6A-L997, or FOTP-60, is a method published in
1997 for measuring fiber or cable length
 Oueruiew Opticetl Fiber Communieations

sgstem standardsrefer to measurement methods for links and networks The-rnajor organizations
and Electronic Engineers
ar; the American National Standards Institute (ANSI), the Institute for Electrical
(IEEE), the ITU-T, and Telcordia Technologies. Of panicular interest for fiber optics system are test
(in number range G.650 and
standards and recorirmendations ffom the ITU-T. wiifrin G
the series the
amplifiers, wavelength multiplexing, optical
fiigfreri the recommendations r€late to fiber cables, optical
,iirpon nerworks (oTN), system reliabiljty and availabilify, *d management and coltrof for passive

optical nerworks (poN). Tli" L.seri"s lru.t

recommendations deal with the ciinstruction,'installation'
rfi?ointenance support,,monitoriiig, and testing of cable and other elements
in the optical fiber outside
Technologies provides a wide range of generic
fionr, ,frur is, tlhe fielded cablelsystem. Teicordia the GR-3120
ieqoirements for telecommU*e[tion network components and systems: For.example'
document describes the necessary
Ciiert, n"qutrements foi'Hatdened Fiber Optic Connectors in the field'
hardened and can be mated
specifications for opticJ connectors that are enl'ironmentally

I 'L

functions
io*prr.r-Uured simulation'and modeling tools that integrate componeni, 1ink, and network
-.u, ,i,uk" the design'process of'complex optical links and networks more efficient, less,expensive, and
personal computers led to the development
i;;";Jr", rt dr"liferatiorr andlncrease in capabilities of
"ripif
oi,,,*v rophirti.ut"d';ii";lafion programs for these for predicting photonic component, link,
lachines
on well-established numerical models
arJr[i;"ri p"*oi**a" behavioi. T[ese software tools are based
anal can simulate factors such as connector losses dueto geometric or position mismatches of fibers,
of coupli*g.gplical pjwer from light sources fibers, behaviors of passive and active
intg ufto can model many
"rn"i"r"i"*
opticat components,,*A lfr" peiformance of complex optical networks.
J[1V j

aet* {eviceg, .1uch optical'filters, waveguide grating

,r;nas-"i rrJ;ir as wavlquiae
.gouplers,
oflsophistiCation:
l

#ays, and optical ao***, to'a'ttgh degree . ' i

1.8.1 SimulationTol"-Gharacteristics . :, , .i:

deqign (CAy)lools can ofler a powerf'ul method toassist in analyzing the'design of an
-o*putir-afand poinls to consider,
oprieJ C.rponent,'ci.Cuii, br irenVork before costly prototypes T",brl, Important
fri*"""i,*"'the approit'meitirjr 'and modeling assumptions made in the software design Since most
telecommunication systems designed with ,ererai decibels of safety margin, approximations for
#i
general necessary
;.ri",*g operating Lehavior that areieasonably accurate,are not only acceptable but in l

to allow uactable computation times' l

the following characteristics:

The theoretical models used'in computer simulations nominally include
j^'"j|i,;;;^j""|i r" ,t all factors which could influence the performance of the component, circuit. il,

"
or neiwork can be apprcpnately evaluated'
. ;;;;;;;'sEOf g+iAete*.u ttrut simulated devices can be inteiconnected with each other to
f;; "ircuits or rietworts. that all possible
. Interfaces tt ut paSs''s"ftrCiiAt information between the constituent components"so

o speed, so that quick

Computational effiGiehcy that allows a trade-off between accuracy and
estimates of systein peifoimance can be made in the early stages of a design'
I The capability to simtilate devices over the desiied spectral bandwidth'
o The ability to simJlate factors such as nonlinear eifects, crosstalk between optical
channels,
distortion in lasers, and dispersion in optical fibers'
 the simulation programs normally have
To enable a user to visualize and simulate a system quickly,
the following fearures:
o graphical icons and a graphical user
The ability to create a system schematic based on a library of
(such as optical fibers' filters'
interface (GUI). The iCons represent various system components
power meters' and spectrum analyzers)'
amplifier9 and instrumentation, 1e.g. data sources,
For example' the user
o The ability for the user to interact *itf, ttt" program durilq.a
simulation"
in order to evaluate their effect'
may want to modify a parameter or some op".-uting condition
when the operating range of interest is
This is especially-i-mpo.i*t in the early ,tug"S of a Jesign
being established.
. A wide range of statistical-analysis, signal-processing,-and.display tools'
o and optical specffa, eye diagrams'
Common airpruy forrrats, including tirie waveforms,ilectrical
and error-rate curves.

1.8.2
Commerciallyavailablesimulationtoolsforlightwaveapplications-":":1"1]."",':i::}:::::
(e'g' lasers' modulators'
;;il#;;.##rg r*g"ages. In these languages, system components
p1?,11i:":ll,T-:::::'"li*,i1":
^-) +^nlo
^7^e2-- ara ronrpccnterl hv 2
il1ilr;ffi;ffi;;i;ff;:;;;;;'*"'"q';'i"'*t".T9
ffit#i##,?i,,"'rr#","J-,;r;^rh;;;uiai.ectio"ur:pricalho *::!:1::r,"i::':"*:T',:*
specifies the values of the owrali""lql'-"::^"^:-t
;ffiffi"r;i;; ; ff#;;;;;;i;;; ,h" userIn addition to using preprogrammed modules, users can
,fr" and io ini"rfu.. characteristics.
programming
"orrrpor,"nt
create their own custom J*i"", with either the underlying software code or the graphical
language.
of a complex component' a
irliig'rr"t a set of glaphical icons, one ian put together a simulation
^r *:-,,+^^ A-o oimnlrr cclecte
k, ;;;"fi'?#;;;il;fu1transmission puth,1'::111"111'1.1.3]i"'-**';^':l,Ti:
'#pili"t and connects them together
;#'i":"Hffi#t,Iffiffi;J;;-;;;;;i; and measurement instruments, ,-- rlru^- rL^ .l^-irn io
when the design is completed'
#tr*;*I;"r;;i;;?ate a model of trre optical transmission system. ^nmnlcterl
, r L--^a^.^^ ^- ^ +^^l bar'
on tool tr
ffi;;;""ffiilr r"prarv
"ro
the simulation can be run using control buttons
----^1^-,
a
pP
^-
nq
and connected together, the complex and challengilg
^-l ^L^ll^-^ian
d,l"rrh;;;;it,;*-f,";; s"te"ted of the electrical and
::3f,:
optical
for the user. This ir";;;';;;;sing rearistic ranges o1 the parameters
in an actual application'
ud.res make sense
components and suu-mJdes. It is imlortant that the-p*u*"t"r
the specification sheets of vendors'
In some cases this may entail examining
1.8.3 ExamPle Programs for Student Use
Several commercial vendors offer various suites of software-based modeling-tool modules for optical
for use across all levels
fiber communication*viili;';; rrr* o"tign and planning tools are intended
comparisons ranging from
of lightwave network ;;;;;t, p"rfor.-""-evaluations, and techno1ogy
links, to entire optical networks'
passive and active .o*porr!nt*'and modules,.to complex ransmission
to offer a wide range of settable options
Familiar measurementTr-r*-*r are built into the software
multidimensional sweeps across a parameter
when displaying data from multiple simulation runs and
modules allow data to be manipulated to mimic any laboratory setup'
;;;.iG#-pio""*rinj system integrators, network operators'
Such tools *" in or"-Uy component and system manuficturers,
planning comparative assessments ofvarious
and access service providlrs foifunctions such as capacity
and analyses of WDM system
technologies, optimization of ffansport and service netwoikt,
syntheses
many universities are using these simulation tools
and link designs, and component designs. In addition,
for both research and teaching purposes'
 Owruteu oJ apdcal F-tber Commurtlmtlorts

Abbreviated versions of several simulation modules may be downloaded for noncommercial

educational use from the websites of the tool vendors. These simplified versions contain predefined
component and link configurations that allow interactive concept demonstrations. Among the numerous
demonstration setups are optical amplifier structures, simple single-wavelength links, and WDM links.
The configurations are fixe4 but the reader has the ability to interactively change the operational parameter
values ofiomponents such as optical fibers, light sources, optical filters, and optical amplifiers. As part
of the results that are possible, images of what would appear on the display screen of a standard instrument,
such as a spectrum analyzer or an oscilloscope, enable the user to see the effects on link performance
when various component values change.
The book websiie (http://www.mhhe.com/keiser/ofc4e) describes current offerings available on vendor
websites of the demonstration modules that the reader may download and learn from.))

* pporr-,nt\{s
(ii)
v trl what are the energies in electron volts
(eV) of light at wavelengths 850, 1310,
1490, and 1550 nm?
(b) Consider a l-ns pulse with a 100-nW
amplitude at each of these wavelengths.
How many photons are in such a pulse at
each wavelength?
A WDM optical transmission system is
designed so that each channel has a spectral
width of 0.8 nm. How many wavelength
channels can be used in the C-band?
1.3 Three sine waves have the following periods:
25 1rs, 250 ns, 125 ps. r$fhat are their
frequencies?
1.4 A sine wave is offset 1/6 of a cycle wittr respect
to time zero. What is its phase in degrees and
in radians?
1.5 Consider two signals that have the same
frequency. When the amplitude of the first
signal is at its maximum, the amplitr'rde of the
second signal is at half its maximum from the
zero level. What is the phase shift between
the two signals?
What is the duration of a bit for each of the
following three signals which have bit rates
of 64 kb/s, 5 Mb/s, and 10 Gb/s?
1,7 (a) Convert the following absolute power
gains to decibel power gains: 10". 0.3. l.
4, 10, 100,500,2".
(b) Convert ttre following deqibel power gains
to absolute power gains: -30 dB, 0 dB, 13
dB, 30 dB. lOn dB.
1.8 (a) Convert the following absolute power
levels to dBm values: 1 PW 1 nW 1 mW,
l0 mW 50 mW.
(b) Convert the following dBm values to
power levels in units of mW: -13 dBm'
-6 dBm,6 dBm, 17 dBm.
A signal travels from pointA to point B. (a)If
pointA and 0.125
the signal power is 1.0 mW at
mW at point B, what is the attenuation in dB?
(D) What is the signal power at point B if the
attenuation is 15 dB?
@ A signal passes through three cascaded
amplifiers, each of which has a 5-dB gain'
What is the total gain in dB? BY what
numerical factor is the signal amplified?
@ A 50-km long optical fiber has a total
of 24 dB.If
attenuation pW of 500 optical
power get launched into the fiber, what is the
output optical power level in dBm and inpW?
*t@ Atransmission linehas abandwidth of 2 MHz.
Ifthe signal-to-noise ratio at the receiving end
is 20 dB, what is the maximum data rate that
this line can support? 4 l* L i'; i 'i=i !r
=
"@ (a) At the lowest TDM level of the digitaf'
service scheme, 24 channels of 64 kb/s
each are multiplexed into a 1.5'14-Mb/s
DSl channel. How much is the overhead
that is added?
(b) The next highermultiplexed level, the DS2
rate, is 6.312 Mb/s. How manY DSI
channels can be accommodated in theDS2
rate and what is the overhead?

(c) If the DS3 rate that is sent over a T3line
is 44,376 Mb/s, how many DSZ channels
can be accommodated on a 73 line and
(d) Using the above results, find how many
DSO channels can be sent over a 73 line.
.-. What is the total added overhead?
(9 ro insert low-speed signals such as 64-kb/s
tr npr.nRENCES
1. S. E. Miller and A. G. Chynoweth, eds, Optical
Fiber Telecommunications, Academic, New
York, 1979. This book and the texts listed in
Refs. 2, 3, and 4 contain dozens of topics in all
areas of optical fiber iechnology presented by
researchers from AI&T Bell Laboratories over
a period of more than twenty Years.
2. S. E. Miller and I. P. Kaminow, eds., Optical
Fiber Telecommunications-Il, Academic, New
York, 1988.
I. P.
-.,
Fibir
Kaminow and T.L Koch , eds., Optical
Telecommunications-Ill, vols. A and B'
Academic, New York, 1997.
4. I. P. Kaminow and T.Li, eds., Optical Fiber
Telecommuni.cations-IV, vols. A and' B,
Academic, New York, 2002.
5. J. M. Senior, Optical Fiber Communications,
Prentice-Hall, Englewood Cliffs. NJ, 2nd ed.,
t992.
6. R. L. Freeman, Fiber-Optic Sysrc;s.f@i
Te lecommunications, Wiley, Hoboken' NJ,
2002.
7. G. P. Agrawal Fiber Optic Communiution.
Systems,Wiley, Ho6oken, NJ, 3rd eA., 2W/
8. R. Ramaswami and K. N. Sivarajan, Optical
Networlcs, Morgan Kaufrnann, San Francisco,
2nd ed.,2OO2.
9. E. Desurvire, Erbium-Doped Fiber Ampkfislpi
Wiley, Hoboken, NJ, 2002.
10. E. Desurvire, D. Bayart, B. Desthieux, and S'
Bigo, Erbium-Doped Fiber Amplifiers: Devices
and System Developmmts,'Wiley, New York,
2002.
11. B. Razavi, Design of Iniegiated Circuits for
Optic al Coitmunicartons, McGraw-Hill, New
York, 2003.
il
I
I
i
i
I
!
I
voice channels into a SONET frame, 84
columns in each SPE are divided into seven
groups of 12 columns. Each such group is called
avirtual tributary.
(a) Whatis the bit rate of such a virtral tibutary?
(b) How many 64-kb/s voice channels can a
virtual tributaly accommodate?
(c) What is the payload efficiencY?
t2. I. A. Buck, Fundamentals of Optical Fibers,
Wiley, New York,2004.
13. K.P. Ho, Phase-Modulated. Optical Communi-
cation System.s, Springer, New York, 2W ',
14. G. Keiser, Opticat Communications Es sehiial*,
McGraw-Hill, New York, 2003.
15. G. Keiser, FTIX Concepts and Applications,
Wiley, Hoboken, NJ, 2(X)6.
t6. A. Yariv and P. Yeh, Photonics: Optical
Electronics in Modern Communications,
Oxford University Press USA, New Yorlf;6tlr
ed.,2OA6. tJ
t7. B. E. A. Saleh and M. C. Teich, Fundamentals
of Photonics, Wiley, Hoboken, NJ, 2nd ed.'
2007.
18. E. Snitzer, "Cylindrical dielectric waveguide
modes," J. Opt. Soc. Amer, vol. 51, pp.49l-
498, May 1961.
19. K. C. Kao and G. A. Hockman, "Dielectric-fibre
surface waveguides for optical frequencies,"
Proceedings IEE, vol. 113, pp. 1151-1158, July
1966.
20. J. Hecht, City of Light,OxfordUniversity Press,
New York, 1999.
2t. Optical Fiber Communications (OFC) Conf. is
cosponsored annually by the Optical Society of
America (OSA), Washington, DC a4*'t{e
Institute of Electrical and Electronic Engiqgeis
(IEEE), New York, NY.
22. European Conference on Optical Fibre Com-
munications (ECOC) is held annually inEurope;
sponsored by various European engineering
organizations.
23. Photonics Wesr held in the US arrd Asia'Pacific
Optical Communications (APOC) held in Asia
are two of a number of annual conferences
sponsored by SPIE, Bellingham, WA, USA-
 Oueruiew of @ticalF'iber
24. Special Issue: "100 years of communication
piogress," IEEE Comm*n. Mag., vol.22,May
1984.
25. A. Iron.Garcia and I.
lWidjaja, Communicqtion
Networks, McGraw-Hill, New York, 2nd ed',
2004.
26. J. G. Proakis, Digital Co-mmunications,
"McGraw-Hill, New'York, 4th €d., 2001.
27. A, B. Carlson, P. Crilly, and J. Rutledge,
Communication Systems, McGraw-Hill,' Burr
Ridge, IL, 4th ed., 2OO2'
28. L. *. Couch fi.,Digital and'Analog Communi-
cation System,s, Prentice Hall, Upper Saddle
River, NJ, 7thed.,2W.
29. F.P. Kapron, D. B. Keck, and R. D' Maurer,
"Radiation losses in glass optical waveguides,"
Appl. Phys. Lett., vol. 17, pp. 423'425,
Nov.1970.
30. W. A. Gambling, "The rise and rise of optical
fibers," IEEE J. Sel. Topics Quantum Electron',
vol. 6, no. 6, pp. 1084-1093 Nov./Dec' 2000'
31. B. St. Arnaud, J. Wu, andB. Kalali,'Customer-
controlled and -managed optical networks," J'
Lightwave Technology,vol. 21, pp' 2804'2810,
Nov. 2003.
32. D. Simeonidou,R.Nejabati,G. Zervas,D' Klonidis,
A. Tzanakaki, M. J. O'Mahony, "Dynamic
optical-netswork architectures and technologies
for existing and emerging grid services," '/'
Lighnvav e kchnology, vol. 23, pp. 1347 -3357'
Oct.2005.
33. N. Taesombut, F. Uyeda, A. A. Chien, L. Smarr,
T. A. DeFanti, P. Papadopoulos, J. Leigh, M'
Ellisman, and J. Orcutt' '"The OptlPuter: High-
performance, QoS-guaranteed network service
ior emerging e-science applications," IEEE
Commun. Mag., vol. 44, pp. 38-45' May 2006'
34. D. H. Rice and G. Kpiser, "Apptcations of fiber
optics to tactical communication systems," IEEE
Co**un. Mag., vol. 23, pp. 46-57, May 1985'
35. ITU-T Recommendation G.Sup39, Optical
Sy stem Desi gn and Engineering Consi'derations,
Feb. 2006.
36. theorem-
A. J. Jeni, "The Shannon sampling
Its various extensions and applications:
A Tutorial Review," Proc. IEEE, vol' 65,
pp. 1565-1596, Nov. 1977.
Conuumiffi
37. H. G. Perros, Connection-Oriented Netvvorks:
SONET/SDH, ATM, MPLS and OPtical
Networlcs,WileY, Hoboken, NJ, 2005.
38. W. Goralski, SONET/SDH,McGraw-Hi11, New
York, 3rd ed., 2003.
39. A. Scavennec and O. Leclerc, "Toward high-
speed 40-Gb/s transponders," Proc. IEEE, vol'
94, pp.986-996, MaY 2006.
40. J.X. bai, C. R. Davidson, M. Nissov, H' Li,
W. T. Anderson, Y. Cai, L. Liu, A. N. Pilipetskii'
D. G. Foursa, W. W. Patterson, P. C. Corbett'
A. J. Lucero, and N. S. Bergano, "Transmission
of 40-Gb/s WDM signals over transoceanic
distance using conventional NZ-DSF with
rrceiver dispersion slope compensation," /'
Lightwave Technology, vol. 24, pp. I9l-200,
Jan. 2006.
4t. E. Le Rouzic and S. Gosselin, "160-Gb/s optical
networking: A prospective techno-economical
analysis," J. Lightwave Technology, vol' 23,
pp. 3024-3033, Oct. 2005'
42. M. S"ho"id"r., S. Vorbeck, R. Leppla, E. Lach,
M. Schmidt, S. B.Papemyi, andK. Sanapi, "Field
transmission of 8 x 170 Gb/s over high-loss
SSMF link using third-order distributed Raman
amplification," l. Lighhuave kchnology,vol' 24'
pp.175-182, Jan. 2006.
43. M. L. Jon"., "Optical networking standards,"
J. Li ghtw av e chnolo gy, v ol. 22, pp. 27 5 -280,
Te
lar 200,4.
M. Telecommunications Sector-International
Telecommunication Union (ITU-T), various
G.600 Series Recommendations for all aspects
of optical fiber communications'
45. Spe-iat issue on "standards activities: Addres-
sing the challenges of next-generation optical
networks," K. Kazi, guest editor, Optical
Networks Magazine, vol. 4, issue 1, Jan'/Feb'
2003.
46. J. Piprek, ed', Optoelectronic Devices:
Advanced Simulation and Analysis, Springer,
New York, 2005.
47 . K.Kawano and T. Kitoh,Introductionto Optical
Waveguide Analysis: Solving Maxwell's
Equation and the Schrddinger Equation,Wrley,
Hoboken, NJ, 2002.
 A. J. Lowery "Phototric simulation t@Is," in
I. Kaminow and'T. I-i; edr.; Optical Fiber
Telecommuniaotiansi IY*B' Systems and
I mp ai rme nt s, Acdernie, 2002.
49. A. J. Lowery, *WDM sysGms sinnilations," in
A. Gumaste and T. Anthony, eds.; DWDM
Nen+tork Desigw and Engi*ering Solutions,
Cisco Press, 2003.
G. Guekos, ed., Photonic Devices for
Telecommunicalions: II w to Model and
Measure, SpringeqNew Yas*, 1998.
51. N. Antoniades, I. Roudas, Q. Ellinaq, aild V.
Grigoryan, Modeling and Computer-Aided
Design of Optical Commwic,atiow Sys;*tas and
ilerworlr,,Springer, New Yorlq 2007.
52. YPlsystems, Inc., Hohdel, Na*, Jersey USA,
wwwvpisystems.com.
53. RSoft Design Group; Inc., Ossioing; New Yortq
USA, www.rsoftdesign.com.
54. Optiwave Systems, [nc., Toronto, Ontario,
Canada, www.optiwave.com.
55. The book website can be f,ouad at
htp ://www. mhha comlkeiser/ofc4e.