Advanced Optical

Communications
Photonics
Transmission
Systems
Course organization
 Teacher:
 Vittorio Curri
 (+39011090)-4178
 [email protected]
 Office: DELEN, old site, third floor
 Timetable
 Monday: 16.00-19.00, room 12D
 Wednesday: 10.00-11.30, room 21A
 To be defined: Simulation laboratory, it will start in november
 Examination
 Written exam
 Virtual lab report
 Oral examination (optional)

Chapter 1: Introduction to optical communication systems 2

 Introduction to
optical communication
systems
 Section 1.1
Introduction
and short history
Photonics
Transmission
Systems
Fields of application
 Main applications: long haul high capacity transport networks
 usually employed to interconnect central offices of telephone
and data carrier
 bit rates higher than 100 Mbit/s
 transmission distances larger than 10 km

 At present, optical fibers are also used to provide connections

 for voice and data to “business” customers
 for home users offering triple play services (voice, data and
TV)

 Other fields of application, not considered in this course, are:

 Sensors
 Medical applications

Chapter 1: Introduction to optical communication systems 5

Photonics
Transmission
Systems
Fiber versus metallic conductor
 “Traditional” cable systems  Fiber optic systems

 Source  Source
 Voltage or current generator  Optical source

 Transmission medium  Transmission medium

 Metallic conductor (usually  Silica glass fiber (it is a
copper) waveguide for the light)

 System performances  System performances

 Maximum transmission  Maximum transmission
distance before distance before
amplification: few kilometers amplification: up to 100 km
 Maximum bit rate over such  Maximum bit rate over such
distances: around 100 Mbit/s distances: 100 Gbit/s and
beyond
Chapter 1: Introduction to optical communication systems 6
Photonics
Transmission
Systems
Basic concepts

…0010111010… …0010111010…
Electric Electric
signal signal

Optical Popt(t)
Photodiode
Source Fiber

Power

time

Chapter 1: Introduction to optical communication systems 7

Photonics
Transmission
Systems
A short history
 Main milestones:
1966: development of glass optical fibers (Corning
laboratories, USA)
1970: development of first semiconductor lasers stable
at room-temperature (AT&T Bell Laboratories, USA)
1978: development of first single-mode lasers at 1550
nm (NTT laboratories, Japan)
1987: development of first Erbium-Doped Fiber
Amplifier (EDFA) (University of Southampton, England)
Around 1990: first attempt of transmission systems with
more than one channel at different wavelength:
Wavelength Division Multiplexing (WDM)
Around 2007: development of ADC allowing DSP-based
coherent receivers for optical communication
Chapter 1: Introduction to optical communication systems 8
Photonics
Transmission
Systems
Capacity and bit-rate evolution
 Research experiments:
1994: 100 Gbit/s (10 Gbit/s per channel)
1996: 1 Tbit/s (10 Gbit/s per channel)
2000: 10 Tbit/s (40 Gbit/s per channel)
2012: Multilevel modulation formats with DSP-baswed
coherent receiver, 32 Gbaud/channel, few modes fibers…

 Commercially available systems:

Bit-rate per channel (wavelength)
1995: 2.5 Gbit/s
1998: 10 Gbit/s
2000: 40 Gbit/s
Today: 100 Gbit/s towards 400 Gbit/s
Chapter 1: Introduction to optical communication systems 9
Photonics
Transmission
Systems
For further reading Historical background

Operation in the Multichannel New Fiber types:

Multimode Low loss low loss window erbium doped  AT&T True Wave
fiber-optic Single mode of 0.2 dB/km @ fiber amplifiers Fiber (NZDSF)
waveguides optical fibers 1550 nm but (EDFAs) @
>5dB/km 1 dB/km @ high dispersion 1550 nm  Corning LEAF
attenuation 1310 nm @ 1550 nm deployed Large Effective
Area Fiber
Early 80s Mid to Late 80s Late 80s to Early 90s
Early 70s Mid 90s Mid 90s
Multimode Development New dispersion
Room Fabry-Perot of single mode shifted fiber Multichannel Dispersion
temperature 1310 nm lasers DFB 1310 nm yields Zero WDM @1550 nm. compensation
GaAs LEDs and and 1550 nm dispersion @ Number of
multimode FP lasers 1550 nm and channels and
Lasers @ 830 0.5 dB/km loss channel spacing
nm @ 1310 nm limited by fiber
four-wave
mixing (FWM)

Chapter 1: Introduction to optical communication systems 10

Photonics
Transmission
Systems
For further reading Early events
The following slides are for further reading only. We will skip them during the course.

 Late ‘700: “transmission” through light signals generated

manually using lamps and mechanical shutters mounted atop
towers
 1880: Alexander Graham Bell invents the Photophone,
Washington
 1950 - 1960: glass rods or fibers are used in mostly in
medicine for illumination and endoscopy
 1945 - 1960: optical communications in free space studied and
then abandoned because of atmospheric impairments
 May 16, 1960: Theodore Maiman demonstrates first laser at
Hughes Research Laboratories in Malibu
 Circa 1960: Bell Telephone Laboratories begin investigating
hollow optical waveguides with regularly spaced lenses for
digital transmission: fibers are discarded because of high losses
(1000 [dB/km])
Chapter 1: Introduction to optical communication systems 11
Photonics
Transmission
Systems
For further reading Early events
 January 1966: Kao says that fiber loss could be reduced below
20 [dB/km] down from 1000 [dB/km]
 Spring 1970: First continuous-wave room-temperature
semiconductor lasers made in St. Petersburg and at Bell Labs
 1970: Corning develops a single-mode fiber with loss of only 17
dB/km at 633 nanometers
 June 1972: multimode fiber with 4 [dB/km] loss at 800 [nm]
 Late 1972: STL modulates diode laser at 1 Gbit/s
 Early 1976: Masaharu Horiguchi (NTT Ibaraki Lab) and Hiroshi
Osanai (Fujikura Cable) make first fibers with low loss: 0.47
dB/km at 1.2 micrometers)
 Spring 1976: Lifetime of best laboratory lasers at Bell Labs
reaches 100,000 hours (10 years) at room temperature

Chapter 1: Introduction to optical communication systems 12

Photonics
Transmission
Systems
For further reading Early events
 May 1977: Bell System starts sending live telephone traffic
through fibers at 45 Mbit/s fiber link in downtown Chicago
 June 29, 1977: Bell Labs announces one-million hours (100-
year) extrapolated lifetime for diode lasers
 Late 1977: AT&T and other telephone companies settle on 850
[nm] gallium arsenide light sources and graded-index multimode
fibers for commercial systems operating at 45 Mbit/s
 September 1978: France Telecom announces plans for fiber to
the home demonstration in Biarritz, connecting 1500 homes in
early 1983
 1978: AT&T, British Post Office and STL commit to developing a
single mode transatlantic fiber cable, using the new 1.3-
micrometer window, to be operational by 1988. By the end of
the year, Bell Labs abandons development of new coaxial cables
for submarine systems
Chapter 1: Introduction to optical communication systems 13
Photonics
Transmission
Systems
For further reading Early events
 Late 1978: NTT Ibaraki lab makes single-mode fiber with record
0.2 dB/km loss at 1.55 micrometers
 1980: Bell Labs publicly commits to single-mode 1.3-micrometer
technology for the first transatlantic fiber-optic cable, TAT-8
 1981: Commercial second-generation systems emerge, operating
at 1.3 micrometers through graded-index fibers.
 1981: British Telecom transmits 140 Mbit/s through 49
kilometers of single-mode fiber at 1.3 micrometers, starts
shifting to single-mode
 1982: MCI leases right of way to install single-mode fiber from
New York to Washington. The system will operate at 400 Mbit/s
at 1.3 micrometers. This starts the shift to single-mode fiber in
America.

Chapter 1: Introduction to optical communication systems 14

Photonics
Transmission
Systems
For further reading Recent History
 1985: Single-mode fiber spreads across America to carry long-
distance telephone signals at 400 Mbit/s and up
 1985: SIP (now Telecom Italia) stops installing coaxial cables in
Italy and shifts to fiber
 Summer 1986: All 1500 homes connected to Biarritz fiber-to-
the-home (FTTH) system
 1986: AT&T sends 1.7 Gbit/s through single-mode fibers
originally installed to carry 400 Mbit/s
 1987: Dave Payne at University of Southampton (UK) develops
erbium-doped fiber amplifier (EDFA) operating at 1.55
micrometers.
 December 1988: TAT-8 begins service, first transatlantic fiber-
optic cable, using 1.3-micrometer lasers and single-mode fibers,
with repeaters. It has a capacity of 7560 x 64 [kbit/s], bit-rate is
155 Mbit/s.

Chapter 1: Introduction to optical communication systems 15

Photonics
Transmission
Systems
For further reading Recent History
 1990: A “Submarine Festoon Network” is developed in Italy, 560
Mbit/s per fiber, 200 km unrepeated links.
 February 1993: Linn Mollenauer of Bell Labs sends 10 Gbit/s
through 20,000 kilometers of fibers using a soliton system with
EDFA amplifiers.
 1994: in the US, CATV migrates to hybrid fiber-coax with EDFAs
 October 1996: The first unrepeated trans-Atlantic link (TAT
12/13) comes to life: 240000 x 64 [Kbit/s], followed by the first
trans-Atlantic link TPC-5 (similar capacity) - single wavelength
per fiber, 5 Gbit/s transmission - longest leg 8000 km

Chapter 1: Introduction to optical communication systems 16

Photonics
Transmission
Systems
For further reading Recent History
 1996 - 2000: thanks to WDM (wavelength division multiplexing)
the performance of optical systems doubles steadily every nine
months
 2000: experimental long-distance systems with 100 channels at
10 Gbit/s per channel (1 Tbit/s) are demonstrated over 10000
km
 2000: experimental long-distance systems with 80 channels at
40 Gbit/s per channel (3.2 Tbit/s) are demonstrated over 500 km
 2000: several trans-Atlantic and trans-Pacific systems are being
built at 160 Gbit/s (16 x 10 Gbit/s) per fiber
 2009: First product based on optical multilevel modulation
format and coherent receiver

Chapter 1: Introduction to optical communication systems 17

 Section 1.2
A quick review of
physics
Photonics
Transmission
Systems
Goals
 Recall some fundamental concepts of electromagnetic
wave propagation

 Properties of an electromagnetic wave:

Frequency and wavelength
Power
Linewidth
Polarization

 Interaction of electromagnetic waves with dielectric

materials

Chapter 1: Introduction to optical communication systems 19

 Subsection 1.2.1
Fundamentals
of electromagnetic wave
propagation
Photonics
Transmission
Systems
Physics constants

Description Value Symbol

Velocity of Light 2.9979 x 10 8 m/s c
Electron Charge 1.6 x 10-19 C q
Planck Constant 6.66 x 10-34 J•s h
Boltzman Constant 1.38 x 10-23 J/K k

Description Value Symbol

Micron 1 x10-6 Meters m
Nanometer 1 x10-9 Meters nm
Kilometer 1 x103 Meters km
Megahertz 1 x 106 Hz MHz
Gigahertz 1 x 109 Hz GHz
Terahertz 1 x 1012 Hz THz

Chapter 1: Introduction to optical communication systems 21

Photonics
Transmission
Systems
Electromagnetic wave propagation
 Propagation in the vacuum in z direction:
 Sinusoidal oscillation in time and in space of the
electromagnetic field
t1 t2 > t1 t3 > t2 t4 > t3
Wave Amplitude
A0
Electric Field

z
Direction of wave
propagation

Wavelength
l0

Chapter 1: Introduction to optical communication systems 22

Photonics
Transmission
Systems
Electromagnetic waves
 The evolution of the electromagnetic field in t and z
can be written as:

E (t , z )  A0 cos(2f 0t   o z  0 )

 where:
A0 = amplitude
f0 = frequency
f0 = initial phase
0 = propagation constant = 2p/l0 (rad/m)

Chapter 1: Introduction to optical communication systems 23

Photonics
Transmission
Systems
Mathematical notation
 The electromagnetic field is
E (t , z )  A0 cos(2f 0t   o z  0 )
 It is usually expressed in terms of “analytic signal”

ˆ
E (t , z )  A0 e j ( 2f 0t  o z 0 )

or in terms of complex envelope

~  j ( o z  0 )
E (t , z )  A0 e
We can notice that the frequency f0
does not appear in the formula for
the complex envelope, and the field
result in a complex baseband signal.
Chapter 1: Introduction to optical communication systems 24
Photonics
Transmission
Systems
Complex envelope
 Conversion formulas:
ˆ  ~
 
E (t , z )  Re E (t , z )  Re E (t , z )e j 2f 0t 
 Signal spectrum:

F  E ( f )
Real signal
spectrum f

 f0 f0

Analytic signal 
F Eˆ ( f ) 
spectrum f

f0

~
F E( f ) 
Complex envelope
spectrum f

Chapter 1: Introduction to optical communication systems 25

Photonics
Transmission
Systems
Frequency and wavelength
 The wavelength is mostly used in the community of
optical communication engineers
 However, since 1998 absolute frequencies have
been standardized for WDM channels
ITU-T G.692 Recommendation

 Conversion formulas:
c
Frequencies – absolute frequencies f 

c
f   2 
Intervals-Bandwidths 
Chapter 1: Introduction to optical communication systems 26
Photonics
Transmission
Systems
Frequency and wavelength

Around a central wavelength of 1550 nm

f  100 GHz    0.8 nm (0.8014 nm)
  1 nm  f  125 GHz (124.78 GHz )

Around a central wavelength of 1310 nm

f  100 GHz    0.57 nm (0.5724 nm)
  1 nm  f  175 GHz (174.6936 GHz )

Chapter 1: Introduction to optical communication systems 27

Photonics
Transmission
Systems
The “optical” electromagnetic spectrum
10

8 UV Visible Infrared
Frequency (1014 Hz)

6
“First” Communication
5 Window (830 nm)
“Second” Communication
Window (1310 nm)
4
“Third” Communication
3 Window (1550 nm)

1
200 400 600 800 1000 1200 1400 1600 1800
Optical Frequency Wavelength (nm)
vs. wavelength f=c/l
Chapter 1: Introduction to optical communication systems 28
Photonics
Transmission
Systems
“Third window”: 1550 nm

200

2.9979e + 005
f [THz ] 
198

196
[nm]
Frequency (THz)

ITU-T reference
194
193.1 THZ

192

190

188 1552.52 nm

186
1500 1510 1520 1530 1540 1550 1560 1570 1580 1590 1600
Wavelength (nm)

Chapter 1: Introduction to optical communication systems 29

 Subsection 1.2.2
Parameters of an
electromagnetic wave
Photonics
Transmission
Systems
Parameters

In addition to frequency (and phase) an

electromagnetic wave is characterized by the
following parameters:

Power

Linewidth

Polarization

Chapter 1: Introduction to optical communication systems 31

Photonics
Transmission
Systems
Power
 Power
Optical power is measured as the total power flowing
through a plane orthogonal to the direction of
propagation
 Usually, the power is expressed in units of
mW, if we use linear units
dBm, if we use logarithmic units
 By definition:
PdBm  10 log10  PmW 
PmW  10 PdBm /10
 Examples of conversion:
1 mW = 0 dBm  - 10 dBm = 0.1 mW = 100 μ W
10 mW = 10 dBm  - 20 dBm = 0.01 mW = 10 μW
100 mW = 20 dBm  - 30 dBm = 0.001 mW = 1 μW

Chapter 1: Introduction to optical communication systems 32

Photonics
Transmission
Systems
Logarithmic units
1
 An example of a very common expression is: Pout   G  Pin
Where: L
Pin and Pout are powers measured in mW
G is a gain and L is an attenuation (both dimensionless
quantities)

 The same expression in logarithmic units becomes:

Pout dBm
  L dB  G dB  Pin dBm

 Notice that logarithmic units are expressed in:

dB when referring to dimensionless quantities
dBm when referring to optical powers

Chapter 1: Introduction to optical communication systems 33

Photonics
Transmission
Systems
Power and complex envelope
 In optics, the following convention is used

 If an electromagnetic field is described in terms of its

complex envelope:
~  j ( o z  0 )
E (t , z )  A0 e
then, its power is given by:
2
P  E  A02


Chapter 1: Introduction to optical communication systems 34

Photonics
Transmission
Systems
Linewidth
 In case of an ideal optical source without modulation,
it emits a sinusoidal signal
 Constant amplitude
 Constant phase
 Constant frequency

 Any real optical source exhibits amplitude and phase

fluctuations
~  j 0 ( t )
E (t )  A0 (t ) e

 These fluctuations are called amplitude noise and phase

noise of the optical source

 These kinds of noise cause the source linewidth that is

evaluatedChapter
considering its FWHM width
1: Introduction to optical communication systems 35
Photonics
Transmission
Systems
Classification of optical sources

 Optical sources (without modulation):

Linewidth smaller than 1 nm
Usually called “coherent” or “monochromatic” sources
Usually based on lasers, which can produce very narrow
spectral lines (few MHz)
Linewidth wider than 1 nm
Usually called “incoherent” sources
Usually based on LEDs (with a typical emission
bandwidth of the order of 10 nm)

Chapter 1: Introduction to optical communication systems 36

Photonics
Transmission
Systems
Polarization
 When considering coherent electromagnetic waves, we
can define the evolution of the electric field in the xy,
the transversal plane with respect to the propagation
direction

E (t , z )  Ax cos(2f 0t   o z   x )  xˆ
 Ay cos(2f 0t   o z   y )  yˆ
ŷ 
 The vector evolves in time
and in z direction
E(t,z)  Polarization properties of
the electromagnetic field
depend on this evolution
x̂

Chapter 1: Introduction to optical communication systems 37

Photonics
Transmission
Systems
Examples of states of polarization

y Elliptic polarization y
45-degree
Linear polarization

x
x

 A laser source emits a linearly polarized light

Chapter 1: Introduction to optical communication systems 38

 Subsection 1.2.3
Modulated signals and
optical spectra
Photonics
Transmission
Systems
Modulated optical signals
 A modulated optical signal can be expressed as an
electromagnetic field

~  j 0 ( t )
Field represented in terms of
E (t )  A0 (t ) e complex envelope, without
considering the dependence on z

 The (instantaneous) power will be given by:

~  j 0 ( t ) ~ 2 2
E (t )  A0 (t ) e  P(t )  E (t )  A0 (t )

Chapter 1: Introduction to optical communication systems 40

Photonics
Transmission
Systems
Modulation and optical spectrum - I
 Most optical transmission systems employ a simple
amplitude modulation
 Therefore, considering an ideal case, the electric field
can be written as:

ˆ
E (t )  A(t ) e  j ( 2f 0t  0 )

 We assume an ideal source without a spurious phase or

frequency modulation (chirp)
 In this case, as we know from signal theory, the spectrum
is equal to the spectrum of A(t), shifted around the
frequency f0

Chapter 1: Introduction to optical communication systems 41

Photonics
Transmission
Systems
Modulation and optical spectrum - II
 Since real optical sources are not ideal, we have:

 j  2f 0t  ( t ) 
E (t )  A(t ) e
 There is a spurious phase modulation Φ(t) depending on
the source characteristics
 This spurious phase/frequency modulation is called
“chirp”
 The resultant spectrum bandwidth occupation can be
larger than the bandwidth of A(t)
This effect is dominant in systems based on LED or directly
modulated laser

Chapter 1: Introduction to optical communication systems 42

Photonics
Transmission
Systems
Modulation and optical spectrum - III

A( f )  F  A(t )
Spectrum of the
baseband useful signal f
B
Spectrum of the optical
signal, with amplitude 2B f
modulation only

f0
Spectrum of the optical signal,  2 B
with amplitude modulation and
spurious phase/frequency
modulation (CHIRP) f
f0
Chapter 1: Introduction to optical communication systems 43
 Subsection 1.2.4
Interaction of light
with dielectric materials
Photonics
Transmission
Systems
Interaction with dielectric materials

 Next slides will be devoted to recall some important

physical principles regarding the interaction of light
(an electromagnetic field) with dielectric materials,
and in particular:
Attenuation
Refraction, reflection and Snell’s law
Dispersion

Chapter 1: Introduction to optical communication systems 45

Photonics
Transmission
Systems
Attenuation
 An electromagnetic field propagating in a dielectric material
attenuates exponentially as a function of distance z

Optical power e-2az

P ( z )  P (0) e 2 z

z
 The attenuation is expressed in dB/km

 Pin 
 dB / Km  10  log10  
P
 out  1Km

 Therefore: P ( z ) dBm  P (0) dBm   z

dB / Km

Chapter 1: Introduction to optical communication systems 46

Photonics
Transmission
Systems
Attenuation – Example
 The best optical fibers, at 1550 nm, exhibit a minimum
attenuation of about 0.2 dB/km
 Example:
 Fiber attenuation: 0.2 dB/km
 Transmission length: 100 km
 Input power: +3 dBm
 Calculate the output power

 dB 
 tot   dB / km  Ltot  0.2   100  km  20 dB
 km 

PTX  3 dBm  2 mW
PRX  PTX   tot  3 dBm  20 dB  17 dBm  0.02 mW

Chapter 1: Introduction to optical communication systems 47

Photonics
Transmission
Systems
Refractive index
 The speed of light in a material depends on the
refractive index n of the material
c
v
n

 Typical values of the refractive index are:

n=1 vacuum (by definition)
n=1.4 to 1.5 glass
n=2.4 lithium niobate (material used for optical
components)

Chapter 1: Introduction to optical communication systems 48

Photonics
Transmission
Systems
Reflection and transmission
 At the interface (plane) between two dielectric materials
with refractive index n1 e n2, reflected and transmitted rays
obey the following two laws:
Re
fle
c ted
m itt ed
s
Tran
qr qt
qi

ent
id
Inc n1 Interface n2

Law of reflection qi = qr

Snell’s law n sinq = n sinq

Chapter 1: Introduction to optical communication systems 49
Photonics
Transmission
Systems
Snell’s law
 We consider: n1  n2

 In this case we have:  t   i

i
n1
Interface between two
dielectric materials
n2
t

Chapter 1: Introduction to optical communication systems 50

Photonics
Transmission
Systems
Total reflection - I
 For angles of incidence qi larger than a critical value

 n2 
 c  arcsin 
 n1 
there is no refracted ray and the light is totally
reflected

i  c
Total reflection is the physical
phenomenon which allow the
n1 propagation of light inside an
optical fiber

n2
Chapter 1: Introduction to optical communication systems 51
Photonics
Transmission
Systems
Total reflection - II
 Example:
n1  1.48, n2  1.46   c  81 

Glass-glass interface: n1 = 1.48, n2 = 1.46

90

80
Transmitted Angle (qt)

70

60

50

40

30  c  81
20

10

0
0 10 20 30 40 50 60 70 80 90

Incident Angle (qi)

Chapter 1: Introduction to optical communication systems 52

Photonics
Transmission
Systems
Refractive index

 The refractive index is (also) a function of:

Wavelength or frequency
This dependence causes “Dispersion” effects
Optical polarization
This dependence causes “Birefringence” effects
Optical power
This dependence causes “Non-linear effects”
Other electromagnetic fields
This dependence causes “Electro-optical effects”

Chapter 1: Introduction to optical communication systems 53

Photonics
Transmission
Systems
Material dispersion
 Let us consider the dependence of the refractive
index on the wavelength

“Rainbow” effect
in a prism
qt(l)
qi
White Light n2(l)>1
n0=1

 The velocity of light varies with wavelength

 This effect is called “material dispersion”

Chapter 1: Introduction to optical communication systems 54

 Section 1.3
Optical fibers
 Subsection 1.3.1
Introduction to optical
fibers
Photonics
Transmission
Systems
Structure of optical fibers
 An optical fiber is a very thin cylinder (usually in glass)
consisting of a central “core” surrounded by a
“cladding”
 The core has a refractive index slightly higher than the
cladding
Glass

Core n1

Cladding n2

Chapter 1: Introduction to optical communication systems 57

Photonics
Transmission
Systems
Physical dimensions
 Glass fibers have the following typical diameters:

 Multi-mode fibers
Cladding: 125 mm
Core: 65 mm

 Single-mode fibers
Cladding: 125 mm
Core: 9 mm

Chapter 1: Introduction to optical communication systems 58

Photonics
Transmission
Systems
Why glass?
 Usually optical fibers are made of glass
(Si02), since it has very low attenuation
values at optical frequencies
 Glass must be as pure as possible
Core is doped with TiO2, GeO2, P2O5, Al2O3
to increase the refractive index

Glass preform

Chapter 1: Introduction to optical communication systems 59

Photonics
Transmission
Systems
Glass fiber attenuation - I

10
Attenuation (dB/km)

Optical fiber

Infrared
1.0 absorption
Rayleigh
scattering

0.1

UV
0.01 absorption

800 1000 1200 1400 1600 1800

First Wavelength (nm) Second Third
Window Window Window
a=1.2 dB/Kkm a=0.4 dB/km a=0.2 dB/km

Chapter 1: Introduction to optical communication systems 60

Photonics
Transmission
Systems
Fiber losses
 The attenuation of glass has a theoretical limit of 0.18
dB/km at 1550 nm
Most of the commercial fibers have now attenuation values
of 0.18 –0.22 dB/km at 1550 nm
On the market, fibers with attenuation 0.16 dB/km are
available
Best fibers produced today (pure-silica core fibers) exhibit
a minimum value of attenuation very close to 0.16 dB/km
 At 1300 nm, typical attenuation values are around 0.4-
0.5 dB/km
 At 850 nm, typical attenuation values are of the order of
2-3 dB/Km
 Moreover, each connector or splice on a fiber introduce
a lumped attenuation of the order of 0.1-0.5 dB
Chapter 1: Introduction to optical communication systems 61
 Subsection 1.3.2
“Step Index” optical
fibers
Photonics
Transmission
Systems
“Step Index” fibers
 Most common optical fibers are “step index”:
They are characterized by an abrupt change in
refractive index between core and cladding

Refractive index profile

n1
n2

Fiber core

r
Fiber cladding

Chapter 1: Introduction to optical communication systems 63

Photonics
Transmission
Systems
“Step Index” fibers
 An important parameter of “step index” fibers, in
addition to core diameter, is the relative refractive
index difference, defined as:
n1  n2

n1
 For a glass fiber, typical values are:
D between 10-2 to 10-3
n1 = 1.5,

Chapter 1: Introduction to optical communication systems 64

Photonics
Transmission
Systems
Propagation of light in optical fibers
 In order to obtain high performances, light must be confined (or
“guided”) within the core when propagating through the fiber
 The total reflection enables light to be confined within the core
 Total reflection depends on
 Angles of incidence (it has to be small enough to allow total
reflection)
 Value of D

Cladding n2

Unguided ray
Air n0 = 1.0 q0 Core n1

Ray guided by the core

Ray not guided by the core

Chapter 1: Introduction to optical communication systems 65

Photonics
Transmission
Systems
Acceptance angle
 If incident rays on the air-core interface are confined within
the “acceptance angle” q0,max, then total reflection occurs
enabling rays to be guided inside the core:
 The acceptance angle is given by the following formula:
 n2 n 2   
1 n1 2
 0,max  sin 1  1 2
  sin 
 n 
 n0   
  0

Fiber acceptance angle

Unguided ray
q0,max

Chapter 1: Introduction to optical communication systems 66

Photonics
Transmission
Systems
Source-fiber coupling
 In order to obtain high coupling efficiency between
the optical source and the fiber, a source has to emit
rays confined within the fiber acceptance angle
Example: if D = 0.02 and n1 = 1.5 (typical values for
multi-mode fibers), we obtain q 0,max ≈ 17˚

Fiber acceptance angle

q0,max ≈ 17˚
Unguided ray

Chapter 1: Introduction to optical communication systems 67

Photonics
Transmission
Systems
Fiber-to-fiber coupling
 In case of fiber-to-fiber coupling, it is very important
the axial alignment
 Axial alignment should be realized with a precision
higher than the core diameter
 For single-mode fibers, it is required a precision of the
order of few m m

Core
diameter Dc

Axial
alignment
error << Dc
Chapter 1: Introduction to optical communication systems 68
Photonics
Transmission
Systems
Fiber connectors

 Single-mode fiber
connectors must have a
mechanical precision of the
order of ≈1 mm
 This causes very high costs
 Typical values of the
insertion loss of a connector
are higher than 0.1 dB
 In order to obtain insertion
loss lower than 0.1 dB,
fiber-to-fiber coupling can
also be realized by means of
fusion splices

Chapter 1: Introduction to optical communication systems 69

 Section 1.4
Light propagation in
optical fibers
 Subsection 1.4.1
Propagation Modes in
optical fibers
Photonics
Transmission
Systems
Propagation Modes in optical fibers
 The electromagnetic field
propagates in an optical fiber
according to “modes” First-order Second-order
 To each “mode” corresponds mode mode

a specific field profile in the I I

cross section, and a specific
wave propagation behavior r r
 A detailed understanding of
the concept of modes
requires the solution to
Maxwell's equations in a
cylindrical geometry
 Analytical solutions can be
expressed by means of Bessel
functions

Chapter 1: Introduction to optical communication systems 72

Photonics
Transmission
Systems
Geometric interpretation
 In this course we consider a simple geometric
interpretation: light can be thought consisting of a number
of "rays“.
 Definition: each mode is associated with a ray that
propagates along the fiber with a specific propagation
angle.
 This is an approach valid for multi-mode fibers.

Mode
#2
z

Mode
#1
Chapter 1: Introduction to optical communication systems 73
Photonics
Transmission
Systems
Propagation modes in optical fibers

 By means of a rigorous mathematical analysis based on

Maxwell’s equations, it is possible to demonstrate that:

1. The number of modes (angles) propagating in a fiber is discrete

and finite

2. Under particular conditions, only a single mode propagates along

the fiber (in the geometrical interpretation, this single mode
propagates parallel to the fiber axis)
In such case we have single-mode fibers

Chapter 1: Introduction to optical communication systems 74

Photonics
Transmission
Systems
Normalized Frequency Parameter: V
 Given:
1. the parameters of a step index fiber (diameter and
refractive indexes of core and cladding)
2. the wavelength of the light
 It is possible to evaluate the parameter V, called
“Normalized Frequency Parameter”:

V  a n n 
1 where:
2 2 2
0 1 2 a = radius of fiber core
n1 = core refractive index
 2  n2 = cladding refractive
  an 2 1
index


l0 = considered
0  wavelength
D = relative refractive
index systems difference
Chapter 1: Introduction to optical communication 75
Photonics
Transmission
Systems
Single-mode condition

 It is possible to demonstrate that an optical fiber is

single-mode if its Normalized Frequency Parameter V
is below a threshold value

It can be shown that:

An optical fiber is a single-mode fiber if: V ≤ 2.405
An optical fiber is a multi-mode fiber if: V > 2.405

Chapter 1: Introduction to optical communication systems 76

Photonics
Transmission
Systems
Example - I
 Let us consider typical values for a step-index
fiber: D = 0.003 and n1 = 1.5
 An optical fiber is single-mode if:
a
V  2.405   3.2943 a is the radius
0 of fiber core

 Therefore, a single-mode fiber has the core

diameter comparable with respect to the
wavelength of the propagating light
 In particular, at 1550 nm, a fiber is single-mode if:
a  3.2943 0  5.1 m
diameter  2a  10.2 m
Chapter 1: Introduction to optical communication systems 77
Photonics
Transmission
Systems
Example - II
 As an alternative, we can say that a fiber is
single-mode if, given their geometrical and
material parameters, the wavelength of the
propagating light is higher than a specific value:
a
0 
3.2943
 This wavelength is called “cut-off wavelength”

 Example: a fiber with diameter = 9 mm, a = 4.5 mm

Cut-off wavelength = 1360 nm

Chapter 1: Introduction to optical communication systems 78

 Subsection 1.4.3
Single-mode fibers vs.
multi-mode fibers
Photonics
Transmission
Systems
Dispersion in single-mode fibers
 We have seen that in a single-mode fiber only one
mode propagates along the fiber, that is a ray parallel
to the fiber axis

 Therefore, intermodal dispersion is null (by definition)

 Performance of single-mode fibers are much better
than multi-mode
Chapter 1: fibers
Introduction to optical communication systems 80
Photonics
Transmission
Systems
Single-mode fibers
 The most common
single-mode fibers
(called SMF “Single SMF
Mode Fiber”, ITU fiber
G.651) are:
Step index
Core diameter: 9 or
10 mm
Cladding diameter:
n
125 mm n  n2  n1  3.5 10 3     2.3 10 3
n1
 SMF are the most used
in telecommunication
networks

Chapter 1: Introduction to optical communication systems 81

Photonics
Transmission
Systems
Single-mode fibers and cables
 Due to mechanical
reasons, fibers are
surrounded by several
layers of plastic material
as a protection
 Then, cables carry
several fiber pairs

Chapter 1: Introduction to optical communication systems 82

Photonics
Transmission
Systems
Single-mode fibers vs multi-mode fibers
 Multi-mode fibers  Single mode
 Even graded-index fibers,  Intermodal dispersion does
have a maximum limit of not occur at all
the order of 1 Gbit/s·km,  Due to a very small core:
due to intermodal they are more sensitive
dispersion to mechanical stresses
 Thanks to a bigger and bends
diameter: it is more difficult to
they are more resistant to connect them
mechanical stresses and  In general: installation is
bends more expensive
they are easier to connect  Single-mode fibers are
 At present, thay are used for transmission
employed in LAN networks distances longer than 1 km
 In the following chapters,
we will mainly focus on
single-mode fiber
Chapter 1: Introduction to optical communication systems 83
Photonics
Transmission
Systems
Chromatic dispersion
 Also in single-mode fibers a dispersion
effect occurs: chromatic dispersion
 Chromatic dispersion: different Transmitted signal spectrum
frequency components of a pulse
propagate through a fiber with
different velocities
 Main cause: refractive index depends f
on frequency f1 f2 Velocity
 Main effect: the received pulse is v2
broadened and causes ISI
 Chromatic dispersion effect is as much v1
higher as the spectrum of the
transmitted signal is larger

Chapter 1: Introduction to optical communication systems 84

Photonics
Dispersion limits for SMF (intensity modulation)
Transmission
Systems

 We list only some numerical results:

2.5 Gbit/s, broad spectral width source (1
nm): transmission distances up to 10 km
2.5 Gbit/s, external modulation: transmission
distances up to 1500 km
10 Gbit/s, broad spectral width source (1
nm): transmission distances up to 3 km
10 Gbit/s, external modulation: transmission
distances up to 60-70 Km

Chapter 1: Introduction to optical communication systems 85

 Section 1.4
Perfomance and optical
multiplexing
 Subsection 1.4.1
Multiplexing techniques
in optical transmission systems
Photonics
Transmission
Systems
Multiplexing techniques
 Multiplexing: simultaneous transmission of multiple
data stream on a shared physical medium
 Multiplexing techniques employed in existing optical
transmission infrastructure are:
Time Division Multiplexing (TDM)
Optical Frequency Division Multiplexing,
normally called “Wavelength Division
Multiplexing” (WDM)

Chapter 1: Introduction to optical communication systems 88

Photonics
Transmission
Systems
Time Division Multiplexing - TDM
 Different data streams are
interleaved (at bit, byte or
word level) in time
Channel 1
t

Channel 2
. t TDM t

MUX
.
.

Channel N TBout
t
TBch Output Digital Stream
Input channels

B ch
 bit rate per channel B Rout  N  B Rch
R

1 T Bch
T ch
 bit duration  ch T Bout 
B
BR N
Chapter 1: Introduction to optical communication systems 89
Photonics
Transmission
Systems
Typical bit-rates
 SONET/SDH bit-rates:

SONET Optical SDH Bit Rate Bit Rate

Carrier approx.
STS-1 OC-1 51.840
Mbit/s
STS-3 OC-3 STM-1 155.520 155 Mbit/s
Mbit/s
STS-12 OC-12 STM-4 622.080 622 Mbit/s
Mbit/s
STS-48 OC-48 STM-16 2488.320 2.5 Gbit/s
Mbit/s
STS-192 OC-192 STM-64 9953.280 10 Gbit/s
Mbit/s
OC-768 STM-256 39813.12 40 Gbit/s
Mbit/s

Chapter 1: Introduction to optical communication systems 90

Photonics
Transmission
Systems
Wavelength Division Multiplexing - WDM

 It is the classical
Optical
Channel 1 spectrum Frequency Division
1 Multiplexing

1 Optical
spectrum
Channel 2
2

.
WDM
. 2
. MUX …
. 
. 1 2 N
.

Channel N An optical WDM multiplexer is

N 
N implemented with a passive device
(filter or coupler)

Chapter 1: Introduction to optical communication systems 91

Photonics
Transmission
Systems
Wavelength Division Multiplexing - WDM
 At the receiver side, a
multi-port filter is used
to separate each WDM Optical
channel: Channel 1 spectrum

 it is called “WDM 1

demux” or “WDM
1
demultiplexer”
Channel 2
Optical WDM 2

spectrum
DE- . . 2

MUX . .
. .
… 
1 2 N Channel N
N 
N

Chapter 1: Introduction to optical communication systems 92

Photonics
Transmission
Systems
WDM: terminology
 WDM transmission is also called “Colored
Transmission”
From the analogy between wavelengths and colors in
the visible range
 Coarse WDM (CWDM): few channels (from 2 to 16)
widely spaced in wavelength (20 nm spacing)

 Dense WDM (DWDM): many channels (up to 80-100)

with narrow spacing (down to 0.4 nm)
Commercial optical systems combine up to 80
wavelengths on a fiber, each one at 10 Gbit/s with a
total capacity of 800 Gbit/s per fiber

Chapter 1: Introduction to optical communication systems 93

Subsection 1.4.2
Photodiodes
Photonics
Transmission
Systems
Photodiodes
 Photodiodes convert the input optical
power in an output current
P(t)
 They are based on semiconductor p-n
junctions
The most common photodiodes are called i(t)
PIN (p-intrinsic-n)
 They are based on the absorption effect
E2
Incident photon

E1
 Each electron traveling through the p-n junction can
generate, with a certain probability, a carrier in the
conduction band
 Therefore the amount of generated current is proportional to
the optical power
Chapter 1: Introduction to optical communication systems 95
Photonics
Transmission
Systems
Ideal photodiode
 For an ideal photodiode we have:

Optical input power Popt Photon energy h f

Probability of photon
Po p t
Number of photons
per unit time
absorption and
carrier generation
d  1
h f (quantum efficiency)

Number of carriers Popt

per unit time d  Carrier charge q
h f

q
Generated current i  d   Popt
h f
Chapter 1: Introduction to optical communication systems 96
Photonics
Transmission
Systems
Example
 Calculate the generated current:
 Received optical power = - 30 dBm at a wavelength of 1550 nm
 Quantum efficiency of the photodiode = 0.95
 Check carefully the units of measure!!
Optical input 30
power Popt  30dBm  10 10
 103 mW  1W

c
38
1
0
m/s


Optical frequency 
f  1
.
9
3
11
04
Hz
1
9
3T
H
z
1
5
5
0
19
0m
q 1.6  1019  A
d   0.95  34 14
 1.18  
h f 6.62  10  1.93  10 W 
Generated
current 
i 
d
q

h

f

P
op
t1
.
1
8
1
03
mA
1
.
18A 
Chapter 1: Introduction to optical communication systems 97
Photonics
Transmission
Systems
Responsivity and noise
 Through photo-detection, an additional noise term
is added to the electrical signal

 Therefore: i (t )  R  Popt (t )  inoise (t )

 where:
d  q
R is the “Responsivity” of the photodiode R
h f
inoise(t) is a random process. It takes into account
noise generated during photo-detection: shot noise

Chapter 1: Introduction to optical communication systems 98

Photonics
Transmission
Systems
Avalanche photodiodes
 Avalanche Photodiode (APD) basic principle:
Unlike PIN photodiodes, each input photon can generate
multiple carriers
Therefore APDs amplify the signal during the detection
process

i (t )  R  GAPD  Popt (t )  i APD

noise (t )

 Given the same input power, APDs generate much

more output current than PINs
 However, APDs generate also more detection noise
than PINs
Chapter 1: Introduction to optical communication systems 99
Photonics
Transmission
Systems
Photodetector model

Photodiode response is not instantaneous,

therefore the photodiode complete scheme is the
following one:
PR  t  Quantum PR  t  Photoelectric
i p (t ) Avalanche i(t )
efficiency effect gain (APD)
i (t )  h  t 

 In PIN photodiodes i(t)=ip(t)

 h(t) accounts for all the filrtering
effects occuring in the photodetection
Chapter 1: Introduction to optical communication systems 100
Photonics
Transmission
Systems
Signal and shot-noise current

The output current can be expressed as the

average current plus a noise component
The noise component is called shot noise

h(t )  i (t )  mi  ish (t )  is  ish

Noiseless signal current

Zero-mean shot noise current. It is a zero-mean

Poisson random process that can be well
approximated with a zero-mean Gaussian
random process
Chapter 1: Introduction to optical communication systems 101
Photonics
Transmission
Systems
Signal and shot-noise current

qd
Responsivity: R ˆ
hf
PR  t  h(t )  i t 

PR0 R PR0

Chapter 1: Introduction to optical communication systems 102

Photonics
Transmission
Systems
Signal and shot-noise current



is  G  R  PR 0  h(t ) dt



 h(t )
2 2 x 2
 sh  q G  R  PR 0  dt


x is the avalanche noise factor

If PIN photodetector is used: G=1, x=1

Chapter 1: Introduction to optical communication systems 103
Photonics
Transmission
Systems
Amplifier(s) and low-pass filter model
i(t)=is+ish

v(t)
Trans-impedance
amplifier
In general, this receiver section is modeled as:
i(t)=is+ish
h(t)
Rt v(t)
vth(t)
(noise)
v(t )  vs (t )  vsh (t )  vth (t )
Chapter 1: Introduction to optical communication systems 104
Subsection 1.4.3
Impairments on
optical links
Photonics
Transmission
Systems
Optical systems performance
At transmission (physical) level, optical
communication systems performance are
limited by following effects:
Noise accumulation
Distortion

Next slides show the different impacts of these

effects on eye diagrams

Chapter 1: Introduction to optical communication systems 106

Photonics
Transmission
Systems
Ideal eye diagram
 If an optical signal propagates ideally through a transmission
line, without any noise or distortion, we should get an ideal
eye diagram:
Pattern representation Eye diagram

Ideal eye
opening

Optimum sampling time

Time Time (2 bit-times)

Chapter 1: Introduction to optical communication systems 107
Photonics
Transmission
Systems
Eye diagram and noise
 Eye diagram for signal with noise: the eye diagram opening
is reduced
 In this case, resultant eye diagrams have a random behavior

Noise on the “1” level

Noise on the “0” level

Chapter 1: Introduction to optical communication systems 108

Photonics
Transmission
Systems
Distortion and ISI
 Distortion: a series of effects that cause a deterministic
(not random) closure of the eye diagram
 Distortion is caused by:
Linear effects (narrow-band filtering, fiber dispersion, etc.)
Non-linear effects (electrical devices saturation, Kerr effect
in optical fiber, etc.)

 Usually distortion causes a broadening of the signal in

time domain
At the transmitter, the time duration of a pulse associated
with a bit is less than (or equal to) a bit-time Tb
At the receiver, distortion tends to broaden the pulse that
can result having a duration larger than a bit-time Tb,
resulting in ISI-effects
Chapter 1: Introduction to optical communication systems 109
Photonics
Transmission
Systems
ISI: Inter-Symbol-Interference

“1” “0” “1”

Transmitted signal
(without noise)
t
-Tb 0 +Tb +2Tb
Received and filtered
signal (without noise)
t
-Tb 0 +Tb +2Tb
t0

Chapter 1: Introduction to optical communication systems 110

Photonics
Transmission
Systems
Eye diagram and distortion

 Impact of distortion  Impact of distortion

and noise

Eye diagram
ISI closure

Chapter 1: Introduction to optical communication systems 111

 Subsection 1.4.4
Performance parameters
in optical communications
Photonics
Transmission
Systems
The Q factor
 As for every digital communications systems, also in
optical communications the unique quality parameter
is the error probability on bits p(e), or, using the
commercially used acronym bit error rate (BER)
 Traditionally, in optical communications, a parameter
with a one-to-one relationship to the BER is used, and
it is called quality or Q-factor. It is related to the BER:

1  1 
BER  erfc Q   Q  2  inverfc 2  BER
2  2 

 And it is usually expressed in dB units as

QdB  20 log10  Q 

Chapter 1: Introduction to optical communication systems 113
 BER vs. Q factor
Photonics
Transmission
Systems

Remember!!! QdB = 20 log10(Q)

0 6 9.5 12 14 15.516.9

1  1 
BER  erfc Q
BER

2  2 

Q (linear)
Chapter 1: Introduction to optical communication systems 114
 Subsection 1.4.5
Performance evaluation of
nonamplified optical links
Photonics
Transmission
Systems
Non-amplified links

PTx P Rx
Tx Channel Rx
BER  F  PRx 

The function F depepends on the modulation format

and on the Tx/Rx structure

Given BERtarget the sensitivity is defined as:

Psens  invF  BER target 

Chapter 1: Introduction to optical communication systems 116
Photonics
Transmission
Systems
IMDD: BER vs. PRx

d1 1

d0 0

1  1 d 1  d 0 
BER  erfc   d 1  d 0  c  PRX
2  2  1   0 

1  c PRX 
BER  erfc  
2 to optical
Chapter 1: Introduction  1   0systems
 2communication  117
Photonics
Non-amplified links: back-to-back performance
Transmission
Systems

BER

BERtarget

P sens
P Rx
Chapter 1: Introduction to optical communication systems 118
Photonics
Non-amplified links: propagation impairments
Transmission
Systems

BER Impairments

BERtarget

Penalty
P Rx
Chapter 1: Introduction to optical communication systems 119
Photonics
Transmission
Systems
Penalty [dB]
Penalty vs. impairment strength

D dB
robustness

Impairment strength
Chapter 1: Introduction to optical communication systems 120
Photonics
Transmission
Systems
Link Budget and Power Budget
 The Link Budget is the balance between the Power
Budget and all the other signal attenuations and
impairments Since it is the
difference between
 The Power Budget is given by: transmitted power and
minimum received

 P sens
power, it is important
PTX dBm dBm
for determining system
performance

 In addition to fiber loss, other signal impairments

can attenuate or distort the signal
 A common approach tries to take into account of
each signal impairment as an “equivalent loss”
that can be added to actual losses

Chapter 1: Introduction to optical communication systems 121

Photonics
Transmission
Systems
System margin
 m is the SYSTEM MARGIN: a safety margin to take into
account other possible impairments (known or
unknown) or component imperfections
 Moreover, it takes into account performance
degradation due to components ageing during the
expected life
 Note: components ageing is sometime already taken
into account in components data sheets

Chapter 1: Introduction to optical communication systems 122

Photonics
Transmission
Systems
Loss Budget

 Signal attenuation and degradation are included in

the Loss Budget (in dB):

 dB / km  L   splices   connectors   bends   propagatio n  

where:
asplices is the loss of all fusion splices, usually 0.10.2
dB per splice
aconnectors is loss of all connectors, usually 0.20.3 dB
each
abends is the total loss due to optical cable bends
apropagation is an equivalent loss due to signal
transmission impairments, such as fiber dispersion
m what’s?
Chapter 1: Introduction to optical communication systems 123
Photonics
Transmission
Systems
Balancing the budget

 In order to satisfy BER requirement specified for

the receiver sensitivity, the following condition
has to be fulfilled:
Power Budget  Loss Budget
Therefore:
PTX dBm  PsensdBm   dB / km  L   splices   connectors   bends   propagation  

 Hence, in realistic systems the maximum

transmission distance, Lmax,eff , is given by:
PTX dBm  PsensdBm   splices   connectors   bends   propagation  
Lmax,eff 
 dB / km
Chapter 1: Introduction to optical communication systems 124
 Subsection 1.4.6
Forward Error Correction
(FEC)
Photonics
Transmission
Systems
Forward Error-Correction
 Typical BER requirements for high capacity
optical communication systems are in the range of
10-9  10-12
 Recently, high capacity optical systems employ
sophisticated transmission technologies such as
Forward Error-Correcting codes (FEC)
 FEC can improve of many dB the optical receiver
sensitivity or, similarly, can reduce of many dB
the OSNR required to obtain a desired BER
 However, this performance comes at the cost of a
bit-rate increment required to transmit
redundancy bits generated by the code

Chapter 1: Introduction to optical communication systems 126

Photonics
Transmission
Systems
How to take into account FEC
Usually, a FEC is characterized by the following
two parameters:
increment of the bit-rate
coding gain (usually given in dB)
How to take into account the coding gain:
For non-amplified systems: adding the coding gain
to the power budget
For systems with optical amplification (single or
multi-span):
calculating the required OSNR
subtracting the coding gain from the required OSNR

Chapter 1: Introduction to optical communication systems 127

Photonics
Transmission
Systems
FEC standards
 At present, there is an international standard for FEC
in optical communications:
Reed Solomon (255,239) code with
Coding gain of 5.8 dB
Overhead of 7%
Bit-rate at physical level has to be increased of 7%
 Moreover there is a further solution:
Concatenated code which utilizes Reed-Solomon
(255,223) and Reed Solomon (255,239)
Coding gain of 7.7 dB
Overhead of 22%
 Other codes are under study

Chapter 1: Introduction to optical communication systems 128

Photonics
Transmission
Systems
Link using FEC
PTx P Rx
Tx Channel Rx
 OH FEC 
R B'  R B   1 
 100


R B'

Precoder Decoder
RB
Digital Digital
data data
Chapter 1: Introduction to optical communication systems 129
Photonics
Transmission
Systems
FEC coding gain

BER

BERtarget on channel

BERtarget on data

FEC
P Rx
FEC GAIN
Chapter 1: Introduction to optical communication systems 130
 Subsection 1.4.7
Performance evaluation of
amplified optical links
Photonics
Transmission
Systems
Amplified links

PTx Noisy OSNR

Tx Rx
Channel
P Rx  Psens

BER  G  OSNR 

The function G depepends on the modulation format

and on the Tx/Rx structure

Chapter 1: Introduction to optical communication systems 132

Photonics
Transmission
Systems
Optical Signal-to-Noise-Ratio - I
 As we have seen previously for non-optically amplified
systems, the performance (sensitivity) is expressed in
function of the average power at the receiver input
 On the contrary, for systems with optical pre-
amplifier, BER is usually expressed in function of a
parameter called OSNR
 The acronym stands for Optical Signal-to-Noise-Ratio
 Therefore, before to examine in depth BER in systems
with an optical pre-amplifier, it is necessary to give an
accurate definition of the OSNR

133
Photonics
Transmission
Systems
Example: IMDD

1 0.98OSNR
BER  e
2

134
Photonics
Transmission
Systems
Amplified links: back-to-back performance

BER

BERtarget

OSNR target
OSNR
135
Photonics
Transmission
Systems
Optical Signal-to-Noise-Ratio -II
 We define the optical signal-to-noise-ratio as:

Pamp
OSNR 
PN
where:
 Pamp is the average power of the signal carrying the information at
the optical pre-amplifier output
 PN is the ASE noise power add by the optical pre-amplifier
 This is captured by a pass-band optical test filter with equivalent
noise bandwidth BN equal to the bit rate RB, centered at the optical
carrier frequency:

BN  RB

136
Photonics
Transmission
Systems
OSNR measurement
OSNR measurements must be taken after
the optical pre-amplifier and before the
receiver optical filter:
Receiver Electrical
optical filter amplifier
PRX (t ) Pamp (t ) clock
recovery
and
RX input decision

photodiode
Optical pre-amplifier
OSNR measurements
must be performed
HERE

 Important: the test filter, used to define and

measure OSNR, has no relationship with the
receiver optical filter
137
Photonics
Transmission
Systems
OSNR measurement
The ideal procedure to measure PN requires the
pre-amplifier output to be disconnected from the
receiver and connected to the test filter followed
by an optical power meter
pass-band
optical test
no input filter with
signal B N = RB
PN

Receiver optical pre- optical power

amplifier meter
(ASE noise)

138