• On the completion of this chapter, you will be able to:

▫ Understand the basic concepts of WDM.

▫ Have a good command of the basic principles, transmission modes, and

composition of WDM.

▫ Know the background and technical characteristics of WDM.

• The linear transmission capability increase of space division multiplexing
(SDM) is implemented by adding optical fibers and transmission devices.

• Time division multiplexing (TDM) evolves from primary rate multiplexing into
quaternary groups in legacy PDH to STM-1, STM-4, STM-16, and STM-64
multiplexing in SDH. The capacity expansion has the following disadvantages:

▫ Disadvantage 1: Services are affected.

▫ Disadvantage 2: The rate upgrade is not flexible.

▫ Disadvantage 3: For a higher-rate TDM device, the cost is high, and the
40 Gbit/s TDM device has reached the rate limit of electronic
components.

• WDM multiplexes optical signals at multiple rates (wavelengths) over one

fiber for transmission. The digital signals carried by these optical signals can
have the same rate and data format or different rates and data formats. New
wavelengths can be deployed on a network to expand the network capacity
based on customer requirements.
• Legend:

▫ Highway: optical fiber

▫ Patrol car: supervisory signal

▫ Refueling station: optical regeneration (amplifier) station

▫ Gray car: client service

▫ Colored car: service on a different channel (wavelength)

▫ Lane: optical wavelength

• If an optical fiber is regarded as a multi-lane highway, a traditional TDM

system uses only one of the lanes. To improve the bit rate of the traditional
system, the speed on the lane must be increased to increase the traffic within
a time unit. However, a dense WDM (DWDM) system uses all these lanes on
the highway to fully utilize the fiber transmission capacity.
• The DWDM technology uses a number of wavelengths as carriers and
transmits the carrier channels over the same optical fiber. Generally, the
DWDM technology is in use with single-mode optical fibers because they
have low loss but provide large bandwidth. In contrast to a single channel
system, DWDM considerably enlarges the network capacity by fully utilizing
optical fiber bandwidth. It features simple expansion, reliable performance,
and especially direct access of multiple services, showing a bright future in
application.
• An OTU converts diverse wavelengths to standard wavelengths specified by
ITU-T using the optical to electrical to optical (O/E/O) scheme. That is, a
positive-intrinsic-negative (PIN) photodiode or an avalanche photodiode
(APD) converts received optical signals to electrical signals, and the electrical
signals are modulated based on the standard wavelength output by the laser
to obtain new optical signals over ITU-T-compliant WDM wavelengths.
• An OMU, located at the transmit end, is a component that has several input
ports and one output port. Each input port receives one optical signal. These
signals are transmitted together through the output port. An ODU, located at
the receive end, has one input port and several output ports and separates
signals at multiple wavelengths.
• An OA (such as BA/LA/PA) amplifies optical signals. An all-optical amplifier
features real-time, high gain, broad bandwidth, on-line, low noise, and low
attenuation. It is an essential component in a next-generation optical fiber
communication system. Erbium-doped fiber amplifiers (EDFAs) and fiber
Raman amplifiers (RFAs) are commonly used in practice. Particularly, EDFAs
have outstanding performance and are extensively used as BAs, LAs, or PAs in
optical fiber communication systems that support long-haul, large-capacity,
and high-speed transmission.
• An OSC is set up to monitor WDM optical transmission systems. ITU-T
recommends the use of 1510 nm wavelength at 2 Mbit/s. The OSC can work
properly at low rates based on high receiver sensibility (better than –48 dBm).
However, it must be dropped before reaching an EDFA and be added after
reaching the EDFA.
• A single-fiber unidirectional WDM system uses two optical fibers. One fiber is
used to transmit optical signals in only one direction, and the other fiber is
used to transmit optical signals in the reverse direction.

• The WDM system can use the huge bandwidth resources of optical fibers, so
that the transmission capacity of each optical fiber can be increased by
several to dozens of times. In a long-haul transmission network, wavelengths
can be added gradually based on the actual service volume to implement
capacity expansion.

• Most WDM systems on the live network use the single-fiber unidirectional
transmission mode.
• A single-fiber bidirectional WDM system uses only one optical fiber to
transmit optical signals in two directions at the same time. The optical signals
in the two directions should use different wavelengths.

• When optical signals need to be amplified to extend the transmission

distance, components such as bidirectional optical fiber amplifiers and optical
circulators must be used. However, the noise figure is relatively poor.
• An integrated DWDM system does not use the wavelength conversion
technology. It requires that the wavelengths for optical signals of
multiplexing terminals comply with the DWDM system specifications.
Different multiplexing terminals send different ITU-T-compliant wavelengths.
In this way, they occupy different channels when connecting to a multiplexer
for wavelength multiplexing.

• You can select different application modes according to the requirements of

the project. In practice, open DWDM and integrated DWDM can be used
together.
• An open DWDM system requires that the optical interfaces of multiplexing
terminals comply with ITU-T standards.

• The DWDM system uses the wavelength conversion technology to convert

the optical signals of multiplexing terminals into specified wavelengths. The
optical signals of different terminals are converted into different ITU-T-
compliant wavelengths, and then are multiplexed.
• Currently, a common optical fiber provides high bandwidth, but its utilization
rate is very low. The DWDM technology increases the transmission capacity of
an optical fiber by several, dozens of, or even hundreds of times the
transmission capacity of a single wavelength. Currently, the highest capacity of
an optical fiber transmission system in commercial use is 48 Tbit/s. Multiple
Raman amplifiers (forward/backward Raman + enhanced Raman) and a remote
optical pumping amplifier (ROPA) can implement a maximum single-span
transmission distance of 400+ km over low-attenuation optical fibers.
• The DWDM system multiplexes and demultiplexes signals by optical
wavelength, regardless of the signal rate and electrical modulation mode. That
is, the multiplexing and demultiplexing are transparent to data. The WDM
system can carry service signals in various formats, such as ATM, IP, and
possible signals in the future. The WDM system transmits signals transparently.
For service-layer signals, each wavelength channel in the WDM system is
similar to a virtual optical fiber.
• During network expansion and development, only optical transmitters and
optical receivers need to be replaced, and optical cables do not need to be
reconstructed. This is an ideal expansion method and also a convenient
method for introducing broadband services (such as CATV, HDTV, and B-
ISDN). In addition, any new service or capacity can be introduced by adding a
wavelength.
• The new communication network using WDM technologies is more simplified
than that using traditional electrical TDM technologies, and has clear network
layers. The grooming of various services can be achieved by only adjusting the
wavelengths of corresponding optical signals. The simplified network structure,
clear network layers, and convenient service grooming bring about high
flexibility, cost-effectiveness, and reliability.
• All-optical switching: Service signals are directly groomed at the optical layer
without optical-to-electrical conversion.
• The 160 wavelengths in the C band are classified into odd and even
wavelengths. The details are as follows:

▫ C-band odd wavelengths: There are 80 odd-numbered wavelengths in

total. The center frequency ranges from 192.150 THz to 196.050 THz
(the center wavelength ranges from 1529.16 nm to 1560.20 nm) and the
frequency spacing is 50 GHz.

▫ C-band even wavelengths: There are 80 even-numbered wavelengths in

total. The center frequency ranges from 192.100 THz to 196.000 THz
(the center wavelength ranges from 1529.55 nm to 1560.61 nm) and the
frequency spacing is 50 GHz.

• The 192 wavelengths in the extended C band are classified into odd and even
wavelengths. The details are as follows:

▫ Extended C-band odd wavelengths: There are 96 odd-numbered

wavelengths in total. The center frequency ranges from 191.350 THz to
196.050 THz (the center wavelength ranges from 1529.16 nm to 1566.72
nm) and the frequency spacing is 50 GHz.

▫ Extended C-band even wavelengths: There are 96 even-numbered

wavelengths in total. The center frequency ranges from 191.300 THz to
196.000 THz (the center wavelength ranges from 1529.55 nm to 1567.13
nm) and the frequency spacing is 50 GHz.
• Compared with a DWDM system, a CWDM system provides a certain number
of wavelengths and a transmission distance within 100 km, which improves
flexibility while reducing system costs. Therefore, the CWDM system is mainly
applied to metro networks. CWDM provides high access bandwidth with low
costs. It is applicable to various popular network structures, such as point-to-
point, Ethernet, and SONET rings, especially for short-distance, high-
bandwidth, and dense access points. For example, CWDM applies to network
communication within buildings or between buildings.
• Answer: ABCD
• This chapter describes the structure, types, and features of optical fibers.

• Through this chapter, you will have a basic understanding of G.652, G.653,
G.654, and G.655 optical fibers as well as concepts such as dispersion.
• Factors, such as the fiber core diameter and refractive index of fiber and
cladding materials, determine optical fiber characteristics.

• The fiber core refractive index n1 must be higher than the cladding refractive
index n2, which is a necessary condition for transmitting optical signals in
optical fibers.

• Optical fibers can be classified into single-mode and multi-mode optical

fibers by transmission mode.

▫ In the case of a single-mode optical fiber, the diameter of the fiber core
is extremely small and generally smaller than 10 μm.

▫ In the case of a multi-mode optical fiber, the diameter of the fiber core
is large and generally about 50 μm.
• When a beam of light radiates to the surface of an object (such as glass)
through the air, the transmission speed slows down, and the transmission
direction changes at the same time. Therefore, the refractive index is
measured by the bending degree of light transmitted from one medium into
another. In general, when a beam of light radiates to the boundary of two
media, it splits into two beams as reflected light and refracted light, as shown
in the preceding figure.
• According to the Snell's law, the incidence angle and refraction angle are
equal (θ1 = θ3), and the relationship between the incidence angle and
refraction angle meets n1sinθ1 = n2sinθ2.
• Light travels at different rates in different media, which seem as if different
transmission media resist transmitting light at different levels. The parameter
used to describe this feature for the medium is named refractive index or
index of refraction. Therefore, if ν is the speed of light in a medium, and с is
the speed of light in vacuum, the refractive index can be determined by using
the following formula: n = с/ν.
• The following figure lists the refractive indexes of different media.
Material Air Water Glass Quartz Diamond

Refractive
1.003 1.33 1.52-1.89 1.43 2.42
Index
• According to the Snell's law, the incidence angle and refraction angle are
equal (θ1 = θ3), and the relationship between the incidence angle and
refraction angle meets n1sinθ1 = n2sinθ2. When the incidence angle is
greater than a critical angle θc, the light is completely reflected on the
contact surface, and no refraction light exists.

• Then, what will happen when light radiates from a medium with large
refractive index (such as the glass) into a medium with small refractive index
(such as the air)?

▫ As shown in the preceding figure, when the incidence angle θ (shown as

the dashed arrow in the figure) reaches a specific value, the refraction
angle (shown as the dashed arrow in the figure) is equal to 90°, and
the light no longer enters the second medium (air in this example). In
this case, the incidence angle is called critical angle θc. If the angle is
further increased to reach θ > θc, all the light will be reflected back to
the incident medium (shown as the solid arrow in the figure). Because
all light is reflected back to the incident medium, this phenomenon is
called total reflection.

• It is easy to understand that the attenuation of an optical fiber can be greatly

reduced when all light is completely reflected in the optical fiber, because the
light is basically transmitted in the fiber core area and does not enter the
cladding.
• Light is a kind of high-frequency electromagnetic wave. Based on the wave
optics and electromagnetic field theory, according to Maxwell's equations, in
an optical fiber, light is transmitted in tens and even hundreds of modes if
the geometry size of the fiber core is much larger than the optical
wavelengths. The number of transmission modes of an optical fiber varies
depending on the fiber diameter. Therefore, step-index fibers or graded-
index fibers can be further classified into single-mode optical fibers and
multi-mode optical fibers by number of transmission modes. A single-mode
optical fiber transmits only one mode of optical signals. The diameter of the
fiber core is small and is generally within the range of 5 μm to 10 μm. A
multi-mode optical fiber transmits multiple modes of optical signals. The
diameter of the fiber core is large and has a typical size of about 50 μm.
• The attenuation of an optical fiber depends on absorption attenuation, scattering
attenuation, and bending attenuation.

• Absorption attenuation: This type of attenuation is caused by the fiber materials,

because the fiber materials absorb a fraction of the light passing through them. It
specially refers to the absorption caused by the presence of impurities, such as
metallic ions and OH- ions.

• Scattering attenuation: Generally, a microscopic bit of change in the fiber

material density and uneven concentration of SiO2, GeO2, and P2O5 ions can
cause uneven refraction in some parts inside an optical fiber. As a result,
scattering occurs. Some light power is scattered outside the optical fiber, and this
is scattering attenuation. Scattering attenuation also occurs due to impurities at
the core-cladding interface, strains, or bubbles generated in the optical fiber
during optical fiber manufacturing. The defect size is much greater than the
optical wavelength, and the scattering attenuation caused by these defects is
irrelevant to the wavelength. The attenuation curve moves upward, which means
that the attenuation increases. However, this type of attenuation is much lower
than the first type of attenuation.

• The attenuation coefficients of single-mode fibers (SMFs) in the 1310 nm and

1550 nm ranges are generally 0.3–0.4 dB/km and 0.17–0.25 dB/km respectively.
According to ITU-T G.652, the attenuation coefficients of optical fibers in the
1310 nm and 1550 nm ranges should be less than 0.5 dB/km and 0.4 dB/km
respectively.
• When signals are carried over optical fibers, different mode components or
frequency components present in the signals have different group velocities.
Therefore, the signals diffuse during transmission, causing signal distortion.
This phenomenon is called dispersion.

• CD: The group velocities vary according to frequencies (or wavelengths) in

the light source. Light with different frequencies arrives at the fiber end with
different time delays. This type of dispersion is known as CD.
• DCF: dispersion compensation fiber
• PMD is the dispersion caused by different speeds of two orthogonal
polarization states of optical signals.

• In an actual optical fiber, the fiber core cross section has a specific ovality due
to the fiber manufacturing process. The uneven thermal expansion
coefficients of the fiber material cause anisotropic stresses on the fiber cross
section, which results in the fiber refractive index anisotropy. Therefore, the
two polarized modes have different propagation speeds, and the delay
difference of group velocities occurs, leading to PMD. PMD is a random value
because it is caused by random factors.

• In an actual WDM system, the system with a rate of 10 Gbit/s or lower is

slightly affected by PMD. For a 40G system, coding technologies can be used
to improve the PMD tolerance.
• The G.652 fiber is the widely used single-mode fiber that has the best 1310
nm performance. It is also called the dispersion-unshifted fiber.

• The G.653 fiber has the best 1550 nm performance and is called dispersion-
shifted fiber (DSF). With the design of the fiber refractive index profile, the
zero-dispersion point is shifted to the 1550 nm window to match the
minimum attenuation window of the fiber. This makes the fiber transmission
at an ultra-high rate and for an ultra-long distance possible.

• The G.654 fiber is the cut-off shifted single-mode fiber. The design of this
fiber aims to decrease the attenuation in the 1550 nm window. The zero-
dispersion point is still around 1310 nm. Therefore, the dispersion in 1550 nm
is high, and can reach 18 ps/(nm·km). In this case, single-longitudinal mode
lasers are required to eliminate the impact of dispersion. The G.654 fiber is
mainly applied to submarine fiber communication that requires long
regeneration sections.

• The G.655 fiber is a non-zero dispersion-shifted single-mode fiber, which is

similar to the G.653 fiber. Therefore, a specific dispersion value is maintained
around 1550 nm, which avoids FWM during DWDM transmission. The fiber is
applicable to DWDM systems.
• Answer: ACD
• If we want to turn DWDM into a commodity, how can we implement it on
hardware? This chapter describes the key technologies and implementation
methods of DWDM, including the light source, optical amplifier, and WDM
components.

• On the completion of this chapter, you will be able to:

▫ Understand the requirements and solutions of DWDM light sources.

▫ Understand the optical amplification technologies of DWDM.

▫ Understand the multiplexing and demultiplexing technologies of

DWDM.

▫ Understand the optical supervisory channel (OSC) of DWDM.

• In a WDM system, the key technologies are mainly classified into four major
technologies: light source technologies, optical multiplexing and
demultiplexing technologies, optical amplification technologies, and
supervisory technology.
• The dispersion tolerance is described as follows: If the dispersion tolerance of
a 10 Gbit/s OTU board is 800 ps/nm, the dispersion coefficient is 20
ps/(nm·km) (considering the dispersion redundancy of the system) and the
dispersion limit (L) is 40 km (L = 800/20 = 40) over the G.652 fiber. That is,
when the transmission distance exceeds 40 km, a DCM must be added for
compensation. Therefore, the larger the dispersion tolerance, the better.
• Modulation is a process in which information of a signal source is processed
and added to a carrier to make the information suitable for channel
transmission. It is a technology in which a carrier changes with a signal.

• Generally, information of a signal source (or information source) includes a

direct current component (DC component) and a frequency component with
a relatively low frequency, and is called a baseband signal.

• A baseband signal cannot be used as a transmission signal. It must be

converted into a signal with a very high frequency relative to the baseband
frequency so that it is suitable for channel transmission. This signal is called
modulated signal, and the baseband signal is called modulation signal.

• Modulation is implemented by changing the amplitude, phase, or frequency

of a high-frequency carrier (or carrier signal of a piece of information), so that
the carrier signal changes with the amplitude of a baseband signal.
Demodulation is a process in which a baseband signal is extracted from a
carrier for a predetermined receiver (or information sink) to process and
understand.
• Direct modulation: It is also called internal modulation, and is used to directly
modulate the light source signal. This modulation mode controls the injection
current of the semiconductor laser to change the intensity of light waves output
by the laser.
▫ Generally, direct modulation on common G.652 fibers can implement
transmission at a rate up to 2.5 Gbit/s over a distance up to 100 km. It
generally applies to CWDM systems.
• EA modulation: It does not directly modulate the light source. In this modulation
mode, an external modulator is used on the output path of the light source to
modulate the light wave. This modulator is equivalent to a switch.
▫ The laser of EA modulation is complex, the loss is large, and the cost is
high. However, EA modulation has a small modulation frequency chirp, and
can be applied to a system with a transmission rate of 2.5 Gbit/s or higher
and a transmission distance of over 300 km. Therefore, in a DWDM system
that uses an optical line amplifier, a laser of the transmitter part is generally
a laser in EA modulation mode.
• Mach-Zehnder (M-Z) modulation: An M-Z modulator splits input light into two
equal channels of signals, which enter its two optical paths. On the two optical
paths, electro-optic material is used and the refractive index of the material
varies with the strength of the externally fed electrical signal. Because a refractive
index change on the optical paths will cause a change in the signal phase, the
combined optical signal at the output end of the two optical paths is an
interference signal with varying strengths. In this manner, information carried on
electrical signals is transformed onto optical signals, achieving modulation based
on the strength of optical signals. The frequency chirp of a separated externally
modulated laser can be equal to zero. Compared with an EA modulated laser, the
separated externally modulated laser has high cost.
• PDM modulation

▫ Step 1: At the transmit end, the PBS splits the laser light into two
perpendicular polarizations (x-pol and y-pol).

• 40G/100G coherent modulation technologies:

▫ ePDM-QPSK, short for enhanced Polarization Division Multiplexing-

Quadrature Phase Shift Keying, is the best solution for 100G WDM
transmission.

▫ ePDM-BPSK, short for enhanced Polarization Division Multiplexing-

Binary Phase Shift Keying, is the solution for 40G ultra-long haul WDM
transmission.
• PDM modulation:
▫ Step 2: After the laser at the transmit end is split into two parts in
polarizations x and y, QPSK/BPSK modulation is performed on the two
polarized light (on the x-pol and y-pol). The PBC combines the
modulated optical signals in polarizations x and y into one optical fiber,
and then the signals are multiplexed and amplified for optical cable
transmission.
• Traditional WDM modulation implements amplitude modulation. Both PDM-
BPSK and PDM-QPSK modulate signal phases.
• The 100G coherent transmission system uses the PDM and QPSK
technologies to reduce the electrical-layer processing rate. For the current
electrical technology, the 40 Gbit/s rate is close to the bottleneck limit.
Higher rate brings about more signal loss, power consumption,
electromagnetic radiation (interference), and impedance match problems,
which cost much financially and technically.
▫ In PDM, one optical signal is divided into two polarizations, and the
signal is then modulated onto the two polarizations. This procedure is
equivalent to splitting a data stream into two parts, reducing the data
rate by half.
▫ For QPSK, a phase equals to two bits, which has the same effect as
dividing data into two parts, and therefore the rate is reduced by half.
▫ For 100G (112 Gbit/s) signals, the actual data baud rate is only 28G
baud (112/2/2 = 28).
• A 40G coherent transmission system can use PDM-BPSK modulation, and a
phase of the BPSK indicates a digital bit. In this way, the decision accuracy and
transmission nonlinear capability are improved.
• QPSK first converts two channels of 28 Gbit/s digital signals (2 bits) into the I-
channel component (real part) and Q-channel component (imaginary part) in
the optical field, and then converts them into a phase signal θ after
calculation using the following formula: s(t) = I x Cosωt – Q x Sinωt = √2
Cos(ωt + θ). This phase signal contains 2 bits. The value of θ can be π/4,
3π/4, 5π/4, and 7π/4, which represent 00, 01, 11, and 10 respectively.
• Coherent detection:

▫ Step 3: A local oscillator (LO) laser that has the same frequency as the
received signal is used to implement coherent processing between the
LO laser signal and received signal to restore the amplitude, phase, and
polarization state information from the received signal.

A coherent light consists of two light waves that meet the coherent condition.

Coherent condition: In the area where they meet, the two light waves
① Have the same oscillation direction.

② Have the same oscillation frequency.

③ Have the same phase or maintain a constant phase relationship with

each other.

Two coherent light waves interfere with each other within the area where they
meet.
• Optical fiber amplifiers (OFAs) can be classified into rare earth ion doped
OFAs and non-linear OFAs. Rare earth ion doped OFAs amplify optical signals
using the simulated radiation mechanism, and non-linear OFAs amplify
optical signals using the non-linear effect of optical fibers. Among OFAs,
EDFAs and RFAs are practical.

• OAs do not need to convert optical signals to electrical signals and then
convert electrical signals back to optical signals. This feature helps form two
advantages over regenerators for OAs.

▫ First, OAs support any bit rates and signal formats, and simply amplify
all received signals. This attribute is generally described as that OAs are
transparent to any bit rates and signal formats.

▫ Second, OAs not only support the amplification of a single signal

wavelength, which is similar to a regenerator, but also support optical
signal amplification within a specific wavelength range. In addition, only
OAs support the TDM and WDM networks with multiple bit rates,
various modulation formats, and different wavelengths.

• Actually, WDM technologies play an important role in optical fiber

communication only after OAs, especially EDFAs, become available. The EDFA
is the most popular OA. It has transformed the theories of WDM and all-
optical network into reality.
• Erbium-doped fibers are the key components of EDFAs and contain Er3+ ions
of a certain concentration. Before illustrating the working principle of EDFAs,
this course introduces the energy level diagram of erbium ions. The outer-
shell electrons of an erbium ion occupy three energy levels (E1, E2, and E3).
E1 is the ground state, E2 the metastable state, and E3 the high-energy state.

▫ When a high-energy pump laser is used to simulate an erbium-doped

fiber, a large number of bound electrons of the erbium ion are
simulated from the ground state to high-energy state (E3).

▫ However, the high-energy state is unstable, and therefore the erbium

ion soon undergoes radiationless decay (photons are not released) and
enters the metastable state E2.

▫ E2 is a metastable energy state, at which a particle can exist for a long

time. The particles simulated by the pump laser continuously assemble
to this state in radiationless transition mode, achieving distribution for
population inversion.

▫ When optical signals with a 1550 nm wavelength traverse the erbium-

doped fiber, particles at the metastable state transit to the ground state
in stimulated radiation mode and generate the same photons as
incident signal photons. In this way, photons in the signal light are
increased, and signals are continuously amplified when traversing the
erbium-doped fiber.
• EDFA: It is a key component of the next-generation optical communication
system. It has the advantages of high gain, high output power, large working
optical bandwidth, independent of polarization, low noise figure, and
independent of system bit rates and data formats. It is an indispensable key
component in a large-capacity DWDM system.

• Erbium-doped fiber: It is one of the most important parts of the EDFA. The fiber
length is about 10 m to 30 m. During manufacturing, erbium is doped to the
quartz fiber core. Therefore, the fiber is called an erbium-doped fiber.

• Isolator (ISO): One isolator is configured before the erbium-doped fiber and
another is configured after the erbium-doped fiber to transmit optical signals in
a single direction.

• Pump light source: 980 nm and 1480 nm pumping supplies are the most
common ones. This is because the 1480 nm pump light source has the highest
laser efficiency, and the 980 nm pump light source has a low noise figure and the
second highest efficiency. The function is to enable the erbium ion to transit
from the low-energy state to the high-energy state.

• Coupler: It combines signal light and pump light and injects them into the
erbium-doped fiber.

• Photoelectric detector: It converts optical/electrical signals. There are two kinds

of commonly used semiconductor photodetectors: PIN photodiode and APD
avalanche photodiode.
• Main advantages of EDFAs:
▫ High coupling efficiency: EDFAs can be easily interconnected with
transmission fibers.
▫ High energy conversion efficiency: The diameter of the erbium-doped
fiber core is smaller than that of the transmission fiber core, and both
the signal light and pump light are transmitted over the erbium-doped
fiber. Therefore, the light energy is concentrated, enabling adequate
interaction between light and gain medium (Er ions). With the erbium-
doped fiber of a proper length, the light energy conversion efficiency is
high.
▫ Good gain stability: EDFAs are insensitive to temperature, and the gain
is independent of polarization. The gain is independent of system bit
rates and signal formats.
• Main disadvantages of EDFAs:
▫ Fixed gain bandwidth range: Due to the differences between energy
levels of Er ions, the operating wavelength range of EDFAs can only be
in the 1550 nm window.
▫ Gain unflatness: The gain bandwidth of EDFAs is large but not flat.
During application in WDM systems, a special technology must be used
to optimize the gain flatness.
▫ Optical surge: The input optical power can be increased quickly using
EDFAs, but the dynamic gain of the EDFAs changes slowly. Therefore,
optical surge may occur in case of a step change of the input signals.
That is, an input optical power peak occurs. Optical surge is more severe
when EDFAs are cascaded. The peak optical power can be several walts,
which may damage O/E converters and optical connector endfaces.
• The gain locking of an EDFA is an important issue. As a WDM system is a
multi-wavelength working system, if the signals of certain wavelengths are
lost, the energy of the WDM system is shifted to the signals that are not lost
due to gain competition, thereby increasing the power of other wavelengths.
At the receive end, the sudden level increase may cause bit errors. In extreme
cases, if seven of the eight wavelengths are lost, all the power will be
concentrated on the remaining wavelength, and the power may reach about
17 dBm. This will result in a strong nonlinear effect or the receive power
overload of the receiver, as well as a large number of bit errors.

• There are many technologies for locking the EDFA gain. The typical method is
to control the gain of a pump light source. The internal monitoring circuit of
the EDFA monitors the ratio of the input power to the output to control the
output of the pump source. When some signals of the input wavelengths are
lost, the input power decreases and the ratio of the output power to the
input power increases. Through the feedback circuit, the output power of the
pump source is reduced to keep the EDFA gain (output/input) unchanged. In
this way, the total output power of the EDFA is reduced, and the output
signal level is stable.
• In a common optical fiber system, the optical power is low, and the optical
fiber has linear transmission characteristics. When the incident optical power
of the light into an optical fiber (non-linear optical medium) is rather high,
the high-energy (short-wavelength) pump light scatters to transfer a small
portion of the incident optical power to the beam whose frequency is down
shifted. The amount of frequency shift depends on the vibration mode of the
medium. This process is called Raman effect. In quantum mechanics, the
process is described as that a photon of the incident wavelength is scattered
to a low-frequency photon, and the molecule performs the transition
between the vibration states. The incident photon is referred to as pump
light, and the low-frequency frequency-shift photon is called Stokes wave.
• Common Raman scattering requires very high laser power. However, in
optical fiber communication, the diameter of a single-mode fiber as a
nonlinear medium is very small (generally less than 10 μm). The single-mode
fiber can be used to restrict the interaction between the high-strength laser
field and the medium to a very small sectional area. This greatly increases the
optical power density of the incident optical field. In a low-loss fiber, the
action between the optical field and the medium can be kept in a long
distance. The energy coupling in the process is very robust, which makes it
possible to utilize SRS in the fiber.
• The quartz fiber has a wide SRS gain spectrum with a wide gain peak around
13 THz that the pump light frequency is down shifted. If a weak signal and a
strong pump light are transmitted in an optical fiber simultaneously, and the
wavelength of the weak signal is within the range of Raman gain bandwidth
of the pump light, the weak signal light can be amplified. An optical amplifier
based on the SRS mechanism is called a Raman amplifier. The gain of the
Raman amplifier is on-off gain, which is the difference of the output power
values when the amplifier is enabled and disabled.
• The Raman amplifier has three prominent characteristics:

▫ The gain wavelength of a Raman amplifier is determined by pump light

wavelengths. Theoretically, optical signals of any wavelength can be
amplified if only the pump light wavelengths are proper. Raman fiber
amplifiers or RFAs can amplify optical signals in the band that EDFAs
cannot amplify. When multiple pump sources are used, RFAs can obtain
a gain bandwidth much larger than that of EDFAs (whose gain
bandwidth is only 80 nm due to limitation of the energy state transition
mechanism). Therefore, the entire low-loss bandwidth range of fibers,
which is from 1270 nm to 1670 nm, can be developed.

▫ The gain medium is the transmission fiber. In this way, RFAs can amplify
optical signals online, which constitutes distributed amplification and
achieves remote pumping and long-haul transmission without
regeneration. RFAs are especially suitable for submarine optical cable
communication where it is inconvenient to deploy repeaters. In addition,
amplified optical signals are distributed along the fiber instead of being
concentrated. Therefore, the optical power is low, which can reduce the
non-linear effect, especially the four-wave mixing (FWM) effect.

▫ The noise figure is low. When being used with EDFAs, RFAs can
significantly lower the system noise figure and extend the transmission
distance.
• According to the position in a DWDM optical transmission network, EDFAs
can be classified into booster amplifiers (BAs), line amplifiers (LAs), and pre-
amplifiers (PAs).

• BA: It is mainly used at the transmit end. It is located after the multiplexer to
increase the power of multiplexed signals and then transmit the signals.
Because the power of multiplexed signals is generally high, there is no high
requirement on the noise figure or gain of the BA; however, it is required that
the output power be relatively high after amplification.

• LA: It is used on the line to periodically compensate for the line transmission
loss. Generally, a small noise figure and a large output optical power of the
LA are required.
• The core components of a WDM system are wavelength division multiplexing
components, which are optical multiplexers and demultiplexers. They are
actually optical filters, and their performance largely determines the
performance of the entire system.

• A multiplexer mainly combines multiple signal wavelengths into one optical

fiber for transmission. A demultiplexer mainly separates multiple signal
wavelengths transmitted over the same optical fiber.

• WDM components determine WDM system performance. The requirements

are as follows: There is plenty of multiplexed channels, the insertion loss is
low, the crosstalk attenuation is large, and the passband range is wide.

• Theoretically, a multiplexer is the same as a demultiplexer. Their only

difference lies in the optical signal input and output directions.
• A TFF comprises tens of alternating layers of high- and low-refractive-index
thin films, which are made of different materials and have varying thickness.
The thickness of each thin film is 1/4 of a wavelength.

• The TFF allows passband transmission of required wavelengths, but blocks

those unwanted wavelengths through a stopband, exhibiting the required
filtering characteristic.

• The main features of a TFF multiplexer are as follows: The component is

small-factor with stable structure, signal passband is flat and independent
from polarization, insertion loss is low, and channel spacing is proper. The
disadvantage is that the number of channels is not large.

• In a WDM system, it is ideal to use TFFs as multiplexers when only 4 to 16

wavelengths need to be multiplexed.
• AWG is a planar waveguide device developed based on optical integration
technologies. Typically, an AWG is fabricated by depositing a thin layer of
silicon dioxide (SiO2) glass on the silicon chip and printing it into required
pattern using photolithography.

• AWGs are commonly used as multiplexers/demultiplexers in DWDM systems.

One outstanding representative of AWGs was developed by NTT (in Japan).
This AWG has small wavelength spacing, supports many channels, and
provides flat passband. It is very suitable for ultra-high speed and large-
capacity WDM systems.
• Supervisory technologies can be classified into optical supervisory
technology and electrical supervisory technology.
• According to ITU-T Recommendations, the OSC of a DWDM system should
be completely independent of the main channel, and the independence of
the OSC from the main channel is fully reflected in the signal flow.

▫ In the transmit direction of an optical terminal multiplexer (OTM) site,

the OSC is connected to the main channel only after the main channel is
multiplexed and amplified. In the receive direction, the OSC is first
separated, and then the system performs pre-amplification and
demultiplexing for the main channel.

▫ In the transmit direction of an optical line amplifier (OLA) site, the OSC
is the last to be connected. In the receive direction of the OLA site, the
OSC is the first to be separated.

• It can be seen that during the entire transmission process, the OSC is not
involved in amplification, but is terminated and regenerated at each site. This
is the opposite of the main channel. The main channel participates in optical
power amplification of the entire process, but is not terminated or
regenerated on the entire line.

• OSC wavelength: 1510 nm

• OSC rate: 2 Mbit/s

• Receiver sensitivity of an OSC: up to –48 dBm

• FA: framing byte

• E1 and E2: orderwire phone

• ALC: automatic level control

• F1, F2, and F3: transparent transmission of data in the same direction

• D1 to D12: DCC bytes

• The SCC board sends supervisory information to an OTU board, and the OTU
board encapsulates signals into its signal transmission unit frame (OTN frame
structure, SDH frame structure, or using the pilot-tone technology).

• The ESC saves the investment for OSC and FIU boards, disregards the
insertion loss of the FIU boards, and improves the optical power budget.
• Answer: D