Optical Amplifiers

3
 Basic Concept
• Optical amplifiers are
insensitive to the bit rate or
signal formats. Thus, a
system using optical
amplifiers can be more easily
upgraded to a higher bit rate,
without replacing the
amplifier.
• Optical amplifiers have fairly
large gain bandwidths and as
a consequence, a single
amplifier can simultaneously
amplify several wavelength
division multiplexing
(WDM) signals, in contrast to
a separate regenerator for
each wavelength.

(a) Amplifier function. (b) Repeater function

Basic Applications of Optical Amplifiers
• Inline optical amplifiers:
It is used as an optical
linear repeater in a long-
haul transmission system
• Pre-amplifier: It is used
as a pre-amplifier to boost
the weak optical signal
before detection.
• Power amplifier: Power
or booster amplifier is
placed immediately after
an optical transmitter to
boost the transmitted
power
Generic Optical Amplifier
 Optical Amplifiers
Two main categories of optical amplifiers are :

• Cavity amplifier/semiconductor optical amplifiers (SOAs):

– Amplification is done by:
• Simulated emission from injected carriers.
• Fiber amplifiers (FA):
– Amplification is done by:
• Stimulated Raman Scattering.
• Doping with rare-earth material such as erbium (Er) or
thulium.
Semiconductor Optical Amplifier
Semiconductor Optical Amplifier
The characteristics of SOA:
• Polarization dependent – requires
polarization maintaining fiber.
• Relatively high gain ~20 dB.
• Output saturation power 5–10 dBm.
• Large Band width.
• Can operate at 800, 1,300, and
1,500 nm wavelength regions.
• Compact and easily integrated with
other devices.
• Can be integrated into arrays.
• High noise figure and crosstalk
levels due to non-linear
phenomenon such as 4-wave
mixing.
 Types of SOA
1. Fabry–Perot amplifier (FPAs).
2. Traveling wave amplifiers (TWAs).

(a) Fabry–Perot amplifier and (b) traveling wave amplifier

• The main difference between the two types of amplifiers is the facet reflectivity
 Fabry–Perot amplifier (FPAs).
Gain and bandwidth of an FPA
Using the standard theory of FP interferometers, the cavity gain of SLA as a function
of signal frequency f is
(1  R1 )(1  R2 )Gs
G( f ) 
(1  R1R2 Gs )2  4 R1R2 Gs sin 2 

  f  fm 
where R1 = input facet reflectivity Gs=the single pass gain
R2 = output facet reflectivity the single pass phase shift  
f
• G reduce to GS when R1=R2=0 fm = cavity resonance frequencies
• G peaks whenever f coincides with f = longitudinal-mode spacing
one of the cavity-resonance frequencies
and drop sharply in between them.
• Amplifier bandwidth is determined by
the sharpness of the cavity resonance
•The ±3dB single longitudinal mode
bandwidth is

2 f  1  R1R2 Gs 
1
BFPA  2( f  f o )  sin  
  2( R R G
1 2 s )1/ 2

 Traveling-wave Amplifier
• TWA operating in the single-pass amplification mode in which the Fabry-Perot
resonance is suppressed by the reduction in facet reflectivity
• antireflection coating is applied to the laser facets - a thin layer of silicon oxide or
silicon nitride
Light output against current characteristic for SLA with different values of facet
reflectivity R

• Antireflection facet coatings have the effect of increasing the lasing current threshold
• In practice SLAs are operated at current far beyond the normal lasing threshold
current
 Crosstalk in SLA
Crosstalk is any distortion of a channel caused by the presence of another channel.
There are two types of crosstalk in SLA:
Interchannel crosstalk - when two wavelengths (channels) enter an SLA, their
nonlinear interference produces new signals at the beat (combination of sums and
differences) frequencies.

Crosstalk in SOA (a) Interchannel crosstalk and (b) Cross saturation

Cross saturation - For a SLA works in the saturated mode (I > IS), the gain
undergoes an opposite change when one channel changes from ON to OFF. This gain
change results in variations in the amplification of another signal because all signals
share the same gain produced by one active medium.
 Doped-fiber Amplifiers
1. An EDFA consists of a 10–30 m length of silica fiber whose core is lightly doped with
ionized atoms of the rare-earth element such as erbium Er3+, ytterbium (Yb), thulium (Tm),
or praseodymium (Pr), while the host fiber material can be standard silica, fluoride glass, or
telluride glass.
2. The major type of fiber
amplifier use in WDM
fiber-optic communication
system is Erbium-doped
fiber amplifiers (EDFAs)
which provide gain near
1.55m wavelength.
3. In EDFA, The fiber is
pumped using a pump
signal from a laser,
typically at a wavelength
of 980 nm or 1480 nm or
both.
4. an optical isolator is used
at the input and/or output
of any amplifier to
prevent back reflections
 3.4.3 EDAF

Erbium fiber = Er3+ doped silica fiber

Pumping wavelength = 980nm or 1480nm
Advantages
(1) Availability of high power pump lasers
(2) All fiber device, polarization independent, ease to
couple, reliable
(3) Simple
(4) Less crosstalk 19
20
Principle of Operation

Stark splitting : an isolated ion of erbium is split

into multiple energy levels.
Each stark splitting level is spread into a band.

21
 Amplification in Erbium-doped Fiber Amplifiers

• optical pumping is obtained by using a pump signal from a laser, typically at a

wavelength of 980 nm or 1480 nm.
• The information signal stimulates transition of the excited Er ions from level 2 to
level 1 and results in radiation of photons with same wavelength, direction, and
phase.
• The energy bands of Er ions in silica fiber gives the EDFA the ability to amplify the
range of wavelengths from 1500 to 1600nm and eliminates the need of fine-tune a
pumping wavelength.
Stark splitting : an isolated ion of erbium is split
into multiple energy levels.
Each stark splitting level is spread into a band.
Thermalization : the erbium ions are distributed in
the various levels within the band.
 Capable of amplifying several wavelengths
simultaneously.
hfc  E2  E1
 32  E3  E2  1  sec
 21  E2  E1  10m sec
atoms stay at E2 longer, 980nm pump is usable
1520nm  f c  1570nm BW  50nm
(c-band from 1530~1565nm)
23
When 980nm pump is used
τ32≈ 1μsec << τ21
We have population inverse between E2 and E1
We can amplify 1530-1570nm signals
When 1480nmpump is used
the absorption from the bottom of E1 to the top of E2
1480nm pump is less efficient
Less population inversion
Higher noise figure
980nm for low noise EDFA
High power 1480nm pump is available
=> High output power and pump can be located
remotely 24
EDFA Gain and Efficiency

• The population at the various

levels within a band is
different; therefore EDFA
gain becomes a function of
the wavelength.
• The gain of EDFAs depends
on:
– erbium-ion concentration,
– amplifier length,
– core radius, and
– pump power
• The Power Conversion
Efficiency (PCE) is
Psignal out  Psignal in
PCE 
Ppump in
26
Gain Flatness

28
29
 Noise in EDFA
• The dominant noise generated
in an EDFA is amplified
spontaneous emission (ASE).

• Signal degradation comes from

beating signals generated at
noise-noise and noise-signal
interference.

• Noise-noise beating can be easily removed by a narrowband filter.

• Noise-signal beatings cannot be filtered because it is within a signal’s bandwidth.
• The amplifier noise figure is 2nsp (G  1)
NFEDFA   2nsp
G
• When a number of optical amplifiers are cascaded in a series, ASE noise of each
stage gets added.
• The system performance is evaluated as Optical Signal Noise Ratio(OSNR): It is
defined as the ratio of the EDFA optical signal output power Pout to the
unpolarized ASE optical noise power PASE. P
OSNR(dB)  10log out

PASE
Raman Amplifiers
(undopped fiber amplifier)
•Raman Amplifiers use SRS effect for amplification.
•The optical signal with short wavelength acts as the "pump"
signal and the energy transfers from it to a modulated weak
signal of a longer wavelength.
•Raman amplification uses common (non-doped) fibers.
•Working around 1310 and 1550 nm
•Require high pump powers
•Broad gain spectrum
 Stimulated Raman Scattering (SRS)
• The interaction between incident
optical signal (photon) and
molecular vibrations gives rise to
Stimulated Raman scattering
(SRS).The Raman threshold
power for single channel optical
Power from lower wavelength channel is system is 32 Aeff
transferred to the higher- wavelength channel PRamanTh 
gR
Where α is attenuation coefficient,
Aeff is the effective core area of an optical fiber
gR is the Raman gain coefficient.

•In WDM systems, the channels at shorter wavelengths will act as pump signals and suffer
from excess loss. On the other hand, the channels at longer wavelengths acting as probe
signals are amplified via SRS
 Fiber Raman Amplifiers
• A fiber Raman amplifier uses stimulating Raman scattering (SRS) occurring in silica fibers
when an intense pump beam propagates through it.
• SRS - the incident pump photon gives up its energy to create another photon of reduced
energy at a lower frequency (inelastic scattering); the remaining energy is absorbed by the
medium in the form of molecular vibrations (optical phonons).
 p   s  op
•The frequency difference,
R   p  s
known as the Stokes shift.

•Because of amorphous nature of glass, the vibrational energy levels of silica molecules
merge together to form a band and allows s differ from p over a wide range (~20THz).
• The Raman amplification exhibits self-phase matching between the pump and signal.
• The pump signal optical wavelengths in Raman fiber amplifiers are typically 500cm
lower than the signal to be amplified, and the pumping signal can propagate in either
direction along the fiber.