OPTICAL NETWORKS

AND
SYSTEM TRANSMISSION
Bus Topology
Ring Topology
Star Topology
T-Coupler
Bypass Scheme using Fiber
Generic of a generic optical T coupler
 T-couplers in Tandem

The coupling Loss is given by

Two tandem passive T couplers
Optical path between Stations
Ring Topology – 3 modes
Topology of linear Bus
 Broadcast-and-Select
WDM Networks
• Broadcast-and-select techniques employing passive
optical stars, buses, or wavelength routers are used
for LAN applications.
• Active optical components form the basis for
constructing wide-area wavelength-routing networks.
• Single-hop refers to broadcast-and-select networks
where information transmitted without O/E
conversions at any intermediate point.
• Intermediate E/O conversion can occur in a multi-hop
broadcast-and-select network.
Architectures for a WDM-based local
network

Fig. Alternate physical architectures for a WDM-based

local network : (a) star, (b) bus.
 Figure shows N sets of transmitters and
receivers being attached to a star coupler or a
passive bus.
• Each transmitter sends its information at
different wavelength.
• All transmissions from various nodes are
combined in a passive star coupler or coupled
onto a bus and the result is sent out to all
receivers.
• Figure below illustrates the concept of multicast
or broadcast for a star network.
• Workstations at nodes 4 and 2 communicate
using l2, whereas a user at node 1 broadcasts
information to workstations at nodes 3 and 5
using l1.
• The same concepts are applicable to bus
structures, although the losses encountered in
the star and bus architectures are different.
A single-hop broadcast-and-select
network
• The WDM setup in Fig. is protocol transparent.
This means that different sets of communicating
nodes can use different information-exchange
rules (protocols) without affecting the other
nodes in the network.
• This is analogous to TDM telephone lines in
which voice, data, or facsimile services are sent
in different time slots without interfering with
each other.
• Figure below shows an example of a four-node
broadcast-and-select multi-hop network where
each node transmits on one set of two fixed
wavelengths and receives on another set of two
fixed wavelengths.
• Stations can send information directly only to
those nodes that have a receiver tuned to one of
the two transmit wavelengths.
• Information destined for other nodes will have
to be routed through intermediate stations.
Fig: Architecture and traffic flow of a multi hop
broadcast-and-select network
Fig: Representation of the fields contained in a data
packet.
• As shown in Fig above, at each intermediate
node, the address header is decoded to examine
the routing information field.
• Using this routing information, the packet is
switched electronically to the specific optical
transmitter.
• The specific optical transmitter will
appropriately direct the packet to the next node
in the logical path towards its final destination.
 Problems in Broadcast-and-Select
Networks
• Two problems in Broadcast-and-Select
networks :
1). More wavelengths are needed as the
number of nodes in the network grows.
2). Without optical booster amplifiers, a large
number of users spread over a wide area
cannot be interconnected.
• Wavelength-routed networks can overcome
these limitations through wavelength-reuse,
wavelength-conversion, optical-switching.
Wavelength-Routed Networks

Fig : Wavelength reuse on a mesh network

 Optical Cross-Connects
• Consider the OXC architecture shown in Fig
below that uses space switching without
wavelength conversion.
• The space switches can be cascaded
electronically controlled optical directional-
coupler elements or semiconductor-optical-
amplifier switching gates.
• Each of the input fibers carries M wavelengths,
any of which could be added or dropped at a
node.
Fig: Optical cross-connect architecture using optical space
switches and no wavelength converters.
• At the input, the arriving signal wavelengths is
amplified and passively divided into N streams
by a power splitter.
• Tunable filters then select individual
wavelengths, which are directed to an optical
space-switching matrix.
• The switch matrix directs the channels either to
one of the eight output lines if it is a through-
traveling signal,
• or to a particular receiver attached to the OXC at
output ports 9 through 12 if it has to be dropped
to a user at that node.
• Signals that are generated locally by a user get
connected electrically via the DXC to an optical
transmitter. The switch matrix directs them to
the appropriate output line.
• The M output lines, each carrying separate
wavelengths, are fed into a wavelength
multiplexer to form a single aggregate output
stream.
• Contentions arise in the architecture shown in
Fig above when channels having the same
wavelength but traveling on different input
fibers enter the OXC and need to be switched
simultaneously to the same output fiber.
• The contentions could be resolved by assigning a fixed
wavelength to each optical path throughout the
network, or by dropping one of the channels and
retransmitting it at another wavelength.
• In the first case, wavelength reuse and network
scalability are reduced.
• In the second case, the add/drop flexibility of the OXC
is lost.
• These blocking characteristics can be eliminated by
using wavelength conversion at any output of the OXC.
Example 12-5 :
• Consider the 4 x 4 OXC shown in Fig.12-21. The OXC
consists of three 2 x 2 switch elements.
• Suppose that l2 on input fiber 1 needs to be switched
to output fiber 2 and that l1 on input fiber 2 needs to
be switched to output fiber 1.
• This is achieved by having the 1st two switch
elements set in the bar-state and the 3rd elements
set in the cross-state, as indicated in Fig below.
• Obviously, without wavelength conversion there
would be wavelength contention at both mux output
ports.
• By using wavelength converters ahead of the
multiplexer, the cross-connected wavelengths can be
converted to non contending wavelengths.
Fig. Example of a simple 4x4 OXC architecture using optical
space switches and wavelength converters.
 Ultra High Capacity WDM Networks
• Two popular approaches are used to achieve increased capacity.
To widen the spectral bandwidth of EDFAs from 30 to 80 nm, by using
broadening techniques.

• Increasing the capacity of a WDM link is to improve the spectral

efficiency of the WDM signals.

• Most of the demonstrations use a rate of 20 Gb/s for each

individual wavelength to avoid non-linear effects.
Examples are,
• (1) A 50-channel WDM system operating at an aggregated 1-Tb/s
rate over a 600 km link.
• (2) A 132-channel WDM system operating at an aggregated 2.6
Tb/s rate over a 120-km/link.
• Repetition rate typically ranges from 2.5 to 10 Gb/s, which corresponds to
the bit rate of the electric data tributaries feeding the system.

• An optical splitter divides the pulse train into N separate streams.

• The pulse streams is 10 Gb/s and N=4, each of these channels is then
individually modulated by an electrical tributary data source at a bit rate
B.

• The modulated outputs are delayed individually by different fractions of

the clock period, and are then interleaved through an optical combiner to
produce an aggregate bitrate of NXB.

• Optical post amplifier and preamplifier are generally included in the link to
compenstate for splitting and attenuation loss.

• At the receiving end, the aggregate pulse stream is demultiplexed into the
original N independent data channels for further signal processing.

• A clock-recovery mechanism operating at the base bit rate B is required at

the receiver to drive and synchronize the demultiplexer.
 SOLITONS
• Group velocity dispersion (GVD) causes most pulses to broaden in time as they
propagate through an fiber.

• A ‘solitons’ are pulses that travel along the fiber without change in shape or
amplitude or velocity.

• Soliton, takes advantage of non-linear effects in silica, particularly self phase

modulation (SPM) resulting from the Kerr non-linearity, to overcome the pulse-
broadening effects if GVD.

• The term “soliton” refers to special kinds of waves that can propagate undistorted
over long distances and remain unaffected after collision with each other.

• In an optical communication system, solitons are very narrow, high intensity optical
pulses that retain their shape through the interaction of balancing pulse dispersion
with the non-linear properties of an optical fiber.

• In an relative effects of SPM and GVD are controlled just right, and the appropriate
pulse shape is chosen, the pulse compression routing from SPM can exactly offset
the pulse broadening effect of GVD.

• Fundamental Solitons- The family of pulse that do not change in shape are called
fundamental solitons.
SOLITONS PULSE
• When a pulse transverse a medium with a
positive GVD parameter β2 for the constituent
frequencies, the leading part of the pulse is
shifted toward a longer wavelength (lower
frequencies) so that the speed in that portion
increases.
• In the trailing half, the frequency arises. So the
speed decreases. This causes the trailing edge
to be further delayed.
• When a narrow high-intensity pulse traverse a medium
with a negative GVD parameter for the constituting
frequencies, GVD counteracts the chirp produced by
SPM.

• GND retards the low frequencies in the front end of the

pulse and advances the high frequencies at the back.

• The result is that the high-intensity sharply peaked

soliton pulse changes neither its shape nor its
spectrum as it travels along the fiber.

• To derive the evolution of the pulse shape required for

sodium transmission, one needs to consider the non-
linear schrodinger (NLS) equation
 Optical CDMA
• The simplest configuration, CDMA achieves multiple access by assigning a unique code to
each user.

• To communicate with another node, user imprint their agreed upon code onto the data. The
receiver can then decode the bit stream by locking onto the code sequence.

• The principle of optical CDMA is based on spread-spectrum techniques.

• The concept is to spread the energy of the optical signal over a frequency band that is much
wider than the minimum bandwidth required to send the information.

• Spreading is done by a code that is independent of the signal itself.

• On optical encoder is used to map each bit of information into the high-rate (longer-code-
length) optical sequence.

• The symbols is the spreading code are called chips.

• The energy density of the transmitted waveform is distributed more or less uniformly over
the entire spread-spectrum bandwidth.

• The set of optical sequences becomes a set of unique ‘address codes or signature sequences’
the individual network users.
The setup consists of N transmitter and receiver
pairs interconnected in a star
• The signature sequence contains six chips. When the data
signal contains 1 data bit, the six-chip sequence is
transmitted, no chips are sent for a 0 data bit.

• Time-domain optical CDMA allows a number of users to

access a network simultaneously, through the use of a
common wavelength.

• Both asynchronous and synchronous optical CDMA

techniques. In synchronous accessing schemes follow rigorous
transmission schedules, the produce more successful
transmission (higher throughputs) than asynchronous
methods where network access is random and collisions
between users can occurs.

• An optical CDMA network is based on the use of a coded

sequence of pulses.
• To send information from node j to node k,
the address code for node k is impressed upon
the data by the encoder at node j.
• At the destination, the receiver differentiates
between codes by means of correlation
detection.
• Each receiver correlates its own address f(n)
with the received signal s(n). The receiver
output r(n) is
• If the received signal arrives at the correct destination, then
s(n)=f(n).

• Equation (5.57) represents an autocorrelation function, if s(n) not

equal to f(n) the equation (5.57) represents a cross-correlation
function.

• For a receiver to be able to distinguish the proper address correctly,

it is necessary to maximize the autocorrelation function and
minimize the cross-correlation function.

• Prime-sequence codes and optical orthogonal codes (OOCs) are the

commonly used spreading sequences in optical CDMA systems.

• An OOC systems the number of simultaneous user an is bounded by

 SONET / SDH

Fig: Basic STS-1 SONET frame

Fig. 12-6: Basic STS-N SONET frame
Fig. 12-7a: 2-fiber UPSR
Fig. 12-7b: 2-fiber UPSR
Fig. 12-8(varied): BLSR architecture
Fig. 12-9: BLSR reconfiguration
Fig. 12-10: Four-fiber BLSR Reconfiguration
Fig. 12-11: Generic SONET network
Fig. 12-12: SONET ADM
Fig. 12-13: Dense WDM deployment
Fig. 12-15: Single-hop network
Important Challenges In Designing Non Linear
Networks Are:

· Transmission of the different wavelength

channel at the highest possible bit rate.

· Transmission over the longest possible

distance with the smallest number of optical
amplifier.

· Network architectures that allow simple

and efficient operation, control and
management.
• The Various Signal Impairment Effect are as Follows
• i. Group Delay Dispersion (GVD), Which limits the bit rate by
temporally spreading a transmitted optical pulse, dispersion
induced pulse spreading can be minimized by WDM networks
operation in a low dispersion window such as 1310 nm and 1550
nm.
• ii. Non-uniform gain across the desired wavelength range of EDFAs
in WDM links.
• iii. Polarization Mode Dispersion (PMD), Which arises from
orthogonal polarization modes traveling at a slightly different
speeds owing to fiber birefringence.
• iv. Reflections from splices and connectors that can cause
instabilities in laser sources. These can eliminated by the use of
optical isolators.
• v. Non-linear inelastic scattering processes, which are interactions
between optical signal and molecular or acoustic vibrations in a
fiber.
• v. Non-linear variations of the refractive index in a silica fiber that
occur because the refractive index is dependent on intensity
changes in the signal.