Chapter 10 - FDDI Backbone Design

1996, BICSI LAN Design Manual - CD-ROM, Issue 1

1
Optical fiber backbones
Overview
The optical fiber used in telecommunications systems is made of extremely pure glassso
pure that the clarity of kilometers of this glass is as transparent as millimeters of the type of
glass used in plate glass windows.
Optical fiber uses light pulses as a means of transmitting signals. The purity of the glass
fibers poses a negligible resistance to the passage of light.
Some of the advantages associated with optical fiber cabling systems include the following:
It has extended reach capabilities. It can easily be used for transmission distances
measured in thousands of meters.
Optical fiber offers extremely high bandwidthenabling the transmission of large
amounts of information. In the case of single-mode optical fiber applications in
campus environments, the bandwidth is considered to be virtually unlimited.
Optical fiber is resistant to electromagnetic interference (EMI) and generates no
radiation of its own. This can be an important factor for high-security installations.
It makes optical fiber difficult to tap intoa tap into an optical fiber cable will result
in a change in the signal strength.
Most current network technologies support optical fiber as a transmission medium,
including Ethernet and Token-ring.
The characteristics associated with optical fiber cables are also discussed in Chapters 1 & 2.
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
2
Selection criteria
Optical fiber is commonly being installed as part of the premises cabling system. Some
considerations that need to be made when selecting optical fiber include the following:
The performance ratings for the optical fiber cable and the connectors to be used. This
includes both attenuation and bandwidth requirements. Attenuation losses must be
considered to ensure the proper operation of the system.
The optical-electronic equipment that will be used. Optical fiber installations require the use
of light sources to convert electronic signals from a computing device into light signals to
inject into the fiber. At the other end, receivers are required to convert the light signals back
into electronic ones. Certain types of optical-electronic equipment may be required,
depending on the type of optical fiber selected.
The environment where the optical fiber will be installed. Cables should be designed for use
in either indoor or outdoor applications. As well, the manner in which the cables will be
distributed must be considered.
Future needs must be anticipated. This may affect the type and number of fibers installed.
Ideally, the cabling should be able to support multiple technologies and adapt to new ones. It
should also be possible to upgrade service, perhaps by installing new optical-electronic
equipment without having to install new cables.
To provide the greatest flexibility and to allow the largest number of applications to benefit
from the installation of optical fiber, ANSI/TIA/EIA-568-A recommendations will be followed
in this discussion.
Selection criteria, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
3
Selecting the optical fiber
ANSI/TIA/EIA-568-A recommends two types of optical fiber for installation in backbone
cabling systems. These are as follows:
62.5/125 m, multimode, graded-index optical fiber.
Single-mode optical fiber.
At this time, single-mode optical fiber is not recommended by TIA/EIA for installation in
horizontal distribution systems.
The installation of composite optical fiber cableswhere both multimode and single-mode
optical fibers are contained in the same cableis acceptable. All fibers must meet the
requirements of the individual optical fiber types. Color coding of fibers becomes even
more important to distinguish the multimode fibers from the single-mode fibersthis
cannot be done otherwise.
Number of fibers
The ANSI/TIA/EIA-568-A standard recommends that optical fiber be installed in groups of
six or twelve fibers for backbone cables.
Most optical fiber applications will require a minimum of two fibersone to transmit
signals and one to receive signals. Many applications also permit redundancya backup
cabling system. This will require an additional two fibers.
Selection criteria, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
4
Backbone cabling distance
The ANSI/TIA/EIA-568-A standard makes the following recommendations regarding the
distances to be used in the backbone cabling system with optical fiber:
62.5/125 m multimode optical fiber.
A maximum of 2000 m (6560 ft) from the main cross-connect to the horizontal
cross-connect.
A maximum of 500 m (1640 ft) from the intermediate cross-connect to the
horizontal cross-connect.
A maximum of 1500 m (4820 ft) from the main cross-connect to the intermediate
cross-connect.
Single-mode optical fiber.
A maximum of 3000 m (9840 ft) from the main cross-connect to the horizontal
cross-connect.
A maximum of 500 m (1640 ft) from the intermediate cross-connect to the
horizontal cross-connect.
A maximum of 2500 m (8200 ft) from the main cross-connect to the intermediate
cross-connect.
Note: Single-mode cable can transmit data for distances in excess of
60 kilometers (37.5 miles) unrepeated; however, distances greater
than 3 kilometers (1.875 miles) are considered outside the scope of
ANSI/TIA/EIA-568-A.
Selection criteria, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
5
Performance criteria
Three performance criteria must be considered operating wavelength, attenuation and
bandwidth. All three vary with the type of optical fiber used. As well, the attenuation and
bandwidth vary with the operating wavelength.
Operating wavelength
The operating wavelength is the wavelength at which the light is being transmitted. Within
the standard acceptable wavelengths, the higher the operating wavelength, the greater the
bandwidth available and the lower the attenuation for the same type of fiber.
Each type of optical fiber is optimized to work at certain wavelength regionsknown as
transmission windows. Both optical fiber types recommended by ANSI/TIA/EIA-568-A are
dual-window fibers. That is, they are optimized to transmit at two different wavelength
regions. These are as follows:
62.5/125 m multimode optical fiber.
Multimode optical fiber is able to support transmission at both 850 nm and 1300 nm
wavelength regions.
Single-mode optical fiber.
Single-mode fiber is able to support transmission at both 1310 nm and 1550 nm
wavelength regions.
Selection criteria, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
6
Attenuation
Attenuation is a measure of how much of the transmitted light is lost. Attenuation losses
must be considered for the cable itself, the connectors being used and any splices that
there are between two cable segments.
Optical fiber cable attenuation is measured as dB/km (decibels of power lost per
kilometer of cable). Therefore, if a cable is rated as having a 3 dB/km attenuation loss, it
is estimated that for every kilometer of this cable installed, the attenuation loss is 3 dB. If
two kilometers of cable are installed there will be a 6 dB loss and if only one-half
kilometer (1640 ft) of cable is installed, the attenuation loss will be 1.5 dB.
The ANSI/TIA/EIA-568-A standard provides the following guidelines regarding cable
attenuation:
62.5/125 m multimode optical fiber.
At an operating wavelength of 850 nm, the maximum acceptable attenuation is
3.75 dB/km.
At an operating wavelength of 1300 nm, the maximum acceptable attenuation is
1.5 dB/km.
Single-mode optical fiber.
At an operating wavelength of either 1310 nm or 1550 nm, the maximum
acceptable attenuation for inside plant cable is 1.0 dB/km and for outside plant
cable is 0.5 dB/km.
Selection criteria, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
7
Optical fiber connectors.
ANSI/TIA/EIA-568-A recommends the use of a connector known as the 568SC
connector. The maximum acceptable optical attenuation for a mated pair of
connectors is 0.75 dB.
Optical fiber splices.
Any optical fiber splice, mechanical or fusion, should not have an optical
attenuation exceeding 0.3 dB.
Selection criteria, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
8
Bandwidth
Bandwidth is a measure of information carrying capacity. It is a measure applied to the
cabling system as a whole. However, it is estimated using the bandwidth measurements of
the optical fiber cable.
Bandwidth is measured as MHzkm (the megahertz of bandwidth available for a one
kilometer transmission). End-to-end bandwidth is usually proportional to distance;
therefore, a cable having a 200 MHzkm bandwidth can transmit a 200 MHz signal for one
kilometer (3280 ft) and can approximately transmit a 100 MHz signal for two kilometers
(6560 ft) or a 400 MHz signal for one-half kilometer (1640 ft).
The ANSI/TIA/EIA-568-A standard provides the following guidelines regarding cable
bandwidth:
62.5/125 m multimode optical fiber.
At an operating wavelength of 850 nm, the minimum acceptable information
transmission capacity is 160 MHzkm.
At an operating wavelength of 1300 nm, the minimum acceptable information
transmission capacity is 500 MHzkm.
Single-mode optical fiber.
At an operating wavelength of 1310 nm or 1550 nm, the information transmission
capacity over a premises cabling system is considered to be virtually unlimited.
Selection criteria, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
9
Optical-electronic equipment
The optical-electronic equipment required includes both a device used to convert
electronic signals into optical pulses and to reconvert the signals back to electronic form.
Light sources
Light sources are used to convert electrical signals into light pulses. They are designed to
operate at specific wavelengths850 or 1300 nm for multimode optical fiber and 1310 or
1550 nm for single-mode optical fiber.
Multimode optical fiber usually makes use of light sources known as Light Emitting Diodes
(LEDs). The lower speed LANs such as Ethernet and Token-ring use such devices
operating at 850 nm. Higher speed LANs such as FDDI and ATM usually require devices
operating at 1300 nm.
Very high-speed systems operating at Gbps speeds with distances exceeding 90 m (295 ft)
usually make use of single-mode optical fiber. The light sources used are lasers. Lasers
are required to provide the very narrow light stream needed to get the light into the very
small core of single-mode optical fiber.
New multimode light sources are being introduced for high-speed LANs, equal to or
greater than 622 Mbps, for in-building cable. These devices include CD lasers and
VSCELs (Vertical Surface Cavity Emitting Lasers) which will provide high-speed
transmission at low cost.
Selection criteria, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
10
A key feature of light sources required by optical fiber network designers is Transmit
Power (T
P
) of the source. The transmit power is a measure of the strength of the signal as
it is launched into the fiber. Transmit power is measured in dBm (decibel as referenced to
one milliwatt) and is usually expressed as a negative value. This value can be obtained
from the manufacturer of the light source.
Receivers
The receiver is required to convert the optical signal back to electrical form. Again,
receivers must be designed for use with a specific type of optical fiber operating at a
specific wavelength.
Receivers have two features important to consider when designing optical fiber networks.
These are the Receiver Sensitivity (R
S
) and Receiver Saturation (R
SAT
).
The receiver sensitivity represents the minimum amount of power (light) that needs to be
delivered to the receiver. Below this level, the receiver cannot see the light.
Receiver saturation represents the point where the light (power) is considered to be too
strong for the receiver. In essence, the light would blind the receiver. If it is determined
that the amount of light arriving at the receiver is too great, an attenuator, or dimmer, must
be installed to reduce the amount of light arriving at the receiver.
Both receiver sensitivity and receiver saturation values are measured in dBm and can be
obtained from the manufacturer of the equipment. Again the dBm measurements are
usually expressed as negative values.
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
11
Link loss budgets
Overview
Link loss budgeting, also known as link engineering, is used to determine if the optical
fiber cabling system being designed will work as expected. It is used to evaluate the total
optical losses of the cabling system.
Link loss budgeting is a three step process, described as follows:
1. Determine the difference between the transmit power of the light source and the
sensitivity of the receiver. This is known as the System Gain or the Dynamic
Operating Range.
System Gain = Transmitter Power - Receiver Sensitivity
2. Determine the total attenuation losses for the cabling system. These are passive
losses attributed to the cable, connectors and splices.
Optical fiber Loss = Length of cable (in km) x Loss per km
Connector Loss = Number of connector pairs x Loss per connector pair
Splice Loss = Number of splices x Loss per splice
Other Loss = Loss due to other components + Margin for component aging
Total Loss = Optical fiber Loss + Connector Loss + Splice Loss + Other Loss
Link loss budgets, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
12
3. Verify that the System Gain exceeds the total passive cabling system losses. The
amount by which the System Gain exceed the total losses represents the safety
margin, which should be sufficient to account for any errors made in estimating
losses or gains.
System Gain > Total Loss
Safety Margin = System Gain - Total Loss
The following example will be used to illustrate how link loss budgets are calculated.
EXAMPLE 10.1: LINK LOSS BUDGETING
An optical fiber campus backbone is to be installed between a main cross-connect and an
intermediate cross-connect. The system will use 62.5/125 m multimode optical fiber and
operate at an 850 nm operating wavelength.
The following information is known about the cabling system:
The total length of optical fiber cable usedincluding all patch cordsbetween the
two points is 1500 m (4920 ft).
There are a total of 6 mated pairs of connectors.
A total of 4 splices will be used.
A 2-dB margin is provided for component aging.
The following information has been provided by the equipment manufacturers:
The transmitter power of the light source, operating at 850 nm is -9dBm.
The sensitivity of the receiver is -28 dBm.
Link loss budgets, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
13
The first step is to calculate the System Gain.
System Gain = Transmitter Power - Receiver Sensitivity
System Gain = -9dBm - (-28 dBm) = 19 dB
The second step is to determine the passive cabling system losses. The loss values
provided by ANSI/TIA/EIA-568-A will be used in calculations. Other components may
require different values to be used.
Optical fiber Loss = Length of cable (in km) x Loss per km
Optical fiber Loss = 1.5 km x 3.75 dB/km = 5.62 dB
Connector Loss = Number of connector pairs x Loss per connector pair
Connector Loss = 6 pairs of connectors x 0.75 dB loss per pair = 4.5 dB
Splice Loss = Number of splices x Loss per splice
Splice Loss = 4 splices x 0.3 dB loss per splice = 1.2 dB
Other Loss = Loss due to other components + Margin for component aging
Other Losses = 2 dB margin for component aging
Total Loss = Optical fiber Loss + Connector Loss + Splice Loss + Other Loss
Total Loss = 5.62 + 4.5 + 1.2 + 2 = 13.31 dB
Link loss budgets, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
14
Finally, Total Loss is compared to the System Gain to see if the system will work.
System Gain > Total Loss ?
19 dB > 13.31 dB
Since the System Gain is greater than the Total Loss, the system will work.
Safety Margin = System Gain - Total Loss
Safety Margin = 19 dB - 13.32 dB = 5.68 dB
The system has a Safety Margin of 5.68 dB.
FIGURE 10.1: LINK LOSS BUDGETING
Graphically, the
various system
losses appear as
shown.
Safety Margin
- 30
- 28
- 26
- 24
- 22
- 20
- 18
- 16
- 14
- 12
- 10
- 8
- 6
- 4
- 2
dBm
System
Gain
Transmitter
Power
X
Receiver
Sensitivity
X
X
X
X
X
Loss due to cable attenuation
Connector loss
Splice loss
Loss due to component aging
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
15
Fiber Distributed Data Interface (FDDI) overview
FDDI defined
FDDI stands for Fiber Distributed Data Interface. It is a second generation LAN technology
examples of first generation LANs are ARCnet, Ethernet and Token-ring.
First generation LANs had several shortcomings, which FDDI was designed to eliminate.
These included the following:
Limited physical area coverage.
Rapid congestion once network traffic increased.
Low reliability and fault tolerance.
Some of the features employed by FDDI to help overcome the shortcomings of other LAN
technologies included the following:
FDDI uses optical fiber cabling as its transmission channel. Therefore, an FDDI
network could span 100 kilometers (62 miles).
FDDI was designed to operate at 100 Mbps. This would permit hundreds of
stations to be linked with acceptable network performance.
An alternate path, to be used in case of primary path failure, was an integral part of
the FDDI specification, greatly increasing network reliability.
FDDI defined, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
16
Since its final adoption by ANSI in 1986, FDDI has proven itself to be an ideal backbone
network technology. It offers the building or campus LAN administrator the ability to link
discrete Ethernets and Token-ring over extended distances at high speeds.
FDDI technology can be used to build a high-speed LAN to the desktop; however, initial high
prices have limited its acceptance except for use with high-speed workstations using high-
bandwidth applications. In 1993, a copper equivalent, TP-PMD, was introduced to allow
Category 5 UTP or STP to be used in the horizontal distribution in hopes of addressing the
installed base and to reduce cost.
As backbone technologies materialize and evolve, users must choose from an ever-greater
selection of technologies. Today, FDDI must compete against other high-speed, shared-
media offerings, such as Fast Ethernet or 100VG-AnyLAN.
As well, high-speed switching technologies such as Asynchronous Transfer Mode (ATM),
switched Ethernet, and even switched FDDI, promise equal or better performance in linking
stations, servers, hubs and networks to one another.
The remainder of this chapter discusses FDDI implementation in a backbone configuration,
where it has been most successfully used since its introduction.
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
17
FDDI networking
Introduction
When standards committees began defining FDDI in 1982, the role for such a network was
clearly in large-scale networking environments. These usually had at least one mainframe
system, and often several mainframe systems, being accessed by thousands of LAN-
connected stations.
As stations continued to increase in power, it became clear that 100 Mbps would someday be
required on even small LANs with a few connected stations.
Today, a single high-performance station operating a high-resolution graphical application can
fill a 100 Mbps channel by itself, by transferring very large files to and from its server.
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
18
Applications
FDDI technology is often used for the following types of applications:
Primary LAN environments.
In such environments, all stations and servers are equipped with FDDI NICs and
are connected to each other, directly or through an FDDI hub.
Primary server environments.
Servers represent the destination for most messages generated by a station. Some
servers can accommodate 1000 simultaneous station links, therefore, the servers
NIC can quickly become a bottleneck. To overcome this problem, the servers are
equipped with 100 Mbps FDDI NICs and are connected to FDDI ports on a hub.
The hub connects to the stations using Ethernet and/or Token-ring LAN technology.
The hub is responsible for performing the frame translation necessary between the
environments.
Backbone LAN environments.
Bridging or routing is used to connect other LAN technologies, such as Ethernet or
Token-ring together over an FDDI network. Where extended distance are
encountered, as on a campus, this method also takes advantage of the long link
lengthsup to 100 kilometers (62 miles)available with FDDI.
Host environments.
FDDI is used to connect mainframe peripherals, such as disk controller units, to
the host at high speeds.
FDDI applications, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
19
Server duplication environments.
In such a configuration, two servers are connected to each other using FDDI, with
one server constantly replicating the others contents. If the first server fails, the
second immediately takes over, maintaining network availability. This is known as
server mirroring.
Because of the extended spans that are possible with FDDI, the two servers can
be placed on different floors or in different buildings, further improving overall
reliability.
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
20
FDDI characteristics
Some of the defining characteristics of FDDI are as follows:
The FDDI specificationsalso known as ANSI (American National Standards
Institute) X3T9.5have also been adopted by ISO (International Organization for
Standardization).
FDDI uses a dual optical fiber ring configuration. One ring is used during normal
operationsthe primary ringand the second ring is available as a backup path in
the event of primary ring failure. Since there is no master node in the network, in
the case of primary ring failure, any node is able to initiate the ring failure recovery
operations.
All stations, servers and/or hubs are connected to each following a ring topology.
Optical fiber cable links to the devices. Most installations make use of 62.5/125 m
multimode optical fiber, but 85/125 m multimode optical fiber and single-mode
optical fiber can also be used.
FDDI uses a token-passing access mechanism similar to Token-ring, making it a
deterministic network technology. A difference between FDDI and low-speed
Token-ring (4 Mbps) is that an FDDI station may release the token immediately
after transmitting its message. This allows for multiple messages to be present at
the same time on the ring.
The specified optical fiber transmitters are LED (Light Emitting Diode)-type, with
an operating wavelength of 1300 nm. Optical fiber receivers are specified as PIN
(Positive Intrinsic Negative)-type diodes.
FDDI characteristics, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
21
A device may be connected to the ring using a single pair of fibersSingle
Attachment Station/Single Attachment Concentrator. Or, a device may be
connected using two pairs of optical fibersDual Attachment Station/Dual
Attachment Concentrator.
The maximum number of dual-attached devices on an FDDI ring is 500. The ring
can have a total span of 100 kilometers (62 miles) and the maximum distance
between two adjacent devices is 2 kilometers (1.2 miles).
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
22
FDDI terminology
Concentrator
A device on an FDDI ring which allows for the connection of multiple stations to the
ring.
Media Interface Connector (MIC)
The specified FDDI connector used to link devices to the network. It uses two
mated ceramic ferrules within a keyed, shrouded assembly.
Single-Attachment Station (SAS)
A station which has a two-fiber link to the primary ring path only, through a
concentrator. Also known as a Class B station.
Single-Attachment Concentrator (SAC)
A concentrator which has a two-fiber link to the primary ring path only.
Dual-Attachment Station (DAS)
A station which has a four-fiber dual link to both the primary and backup ring paths.
Also known as a Class A station.
Dual-Attachment Concentrator (DAC)
A concentrator which has a four-fiber dual link to both the primary and backup ring
paths.
FDDI terminology, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
23
Physical Medium Dependent (PMD) sublayer
The FDDI layer which specifies optical power output and receiver sensitivity, cable
specifications, the MIC interface and optical bypass specifications.
Physical (PHY) sublayer
The FDDI sublayer responsible for transferring network trafficencoding, decoding
and clockingto and from an FDDI device interface.
Media Access Control (MAC) sublayer
The FDDI sublayer responsible for packet framing, addressing and scheduling
frames for transmission over the ring.
Station Management (SMT) specifications
The specifications which define how an FDDI station can be monitored by a
network management service.
FDDI terminology, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
24
FIGURE 10. 2:FDDI CONFIGURATION
Primary ring
DAS
DAS
DAS
DAS = Dual-Attachment Station (Class A node)
SAS = Single-Attachment Station (Class B node)
SAS
Concentrator
Secondary ring
Counter-rotating
rings
Dual
Attachment
Single
Attachment
Primary ring attachment
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
25
FDDI functionality
An FDDI network is made up of two separate pairs of optical fiber ringseach pair
comprises a transmit fiber and a receive fiber. During normal operations, one ring is active
while the other acts in a standby
mode, ready to operate if the
primary ring fails, either
due to a cable fault or a
device fault.
FIGURE 10.3:
DUAL RING
OPERATION
Primary ring
Secondary (backup) ring
FDDI functionality, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
26
If there is a cable fault, such as a break, the devices on either side of the break detect the
fault and wrap their signals onto the secondary ring, which begins carrying signals in the
opposite direction. Thus, a ring is again created, one which bypasses the fault.
FIGURE 10.4:
FDDI RECOVERY
FROM A CABLE BREAK
Cable break
Primary ring
Secondary (backup) ring
becomes active
Signal wraps onto secondary ring
Signal wraps onto
primary ring
FDDI functionality, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
27
The same mechanism occurs in the event of a station faulteven when the station is turned
off by the user. To avoid ring reconfiguration in such cases, devices called optical bypass
units are used. These serve to complete the optical fiber connection through the primary ring
when station power is lost.
FIGURE 10.5:
FDDI RECOVERY
FROM A STATION
FAILURE
Primary ring
Secondary (backup) ring
Optical bypass unit allows signal to
continue to travel over the primary path
FDDI functionality, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
28
Once a faulty cable segment or device has been repaired, FDDI automatically recognizes the
correction and restores the ring to its previous status, with a primary and a standby ring.
FIGURE 10.6:
REPAIRED FDDI RING
Primary ring
Secondary (backup) ring
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
29
FDDI operation
A station with a message to transmit must have possession of a token before it is able to
transmit. Once the station has the token, it creates and transmits a frame. It also releases
the token immediately so that another station is able to transmit.
Stations between the sending device and the destination device have the responsibility of
repeating the frames. At the destination device, a copy of the frame is made and the frame
is repeated. It is the responsibility of the sending device to remove the frame from the ring.
The following example is used to illustrate FDDI
transmission.
EXAMPLE 10.2: FDDI TRANSMISSION
In this example, there are six
devices connected to an FDDI
network. Station A has a
message to transmit to
Station F and at the same
time Station D has a
message to transmit to
Station B.
Both Stations A and D wait for an
available token so they may begin sending
their respective messages. Station A is the first
to gain access to a token (T1) and removes it from
the channel.
Station A
Station F
Station E
Station D
Station C
Station B
T1
FDDI operation, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
30
Station A uses the
token (T1) to begin
transmitting its
message (Frame A).
A new token (T2) is
added to the end of
Frame A by
Station A.
Station A
Station F
Station E
Station D
Station C
Station B
T2 Frame A
FDDI operation, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
31
Station B receives Frame A and since the frame is not for Station B, it regenerates the
frame and passes it to Station C. Station C is also not the destination for the frame so it
also regenerates the frame and passes it to Station D.
FDDI operation, continued
Station A
Station F
Station E
Station D
Station C
Station B
T2 Frame A
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
32
Station D is not the destination for Frame A but it does have a message to send. Station D
removes the token (T2) from the end of the frame, regenerates Frame A and passes it to
Station E.
Station A
Station F
Station E
Station D
Station C
Station B
T2
Frame A
FDDI operation, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
33
Since Station D is in possession of a token (T2) it is able to transmit its message
(Frame D). It appends a new token (T3) to the end of Frame D and puts it onto the
transmission channel. Station E is not the destination for either Frame A or Frame D so
they are both regenerated and passed to Station F.
FDDI operation, continued
Station A
Station F
Station E
Station D
Station C
Station B
T3 Frame D
Frame A
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
34
Station F is the destination of Frame A. This station makes a copy of the frame for itself
and then passes the frame to the next stationStation A. Station F is not the destination
for Frame D so it is regenerated and passed to Station A.
FDDI operation, continued
Station A
Station F
Station E
Station D
Station C
Station B
T3 Frame D
Frame A - copy
Frame A - original
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
35
Station A was the device sending Frame A so it now removes the frame from the channel.
This station is not the destination for Frame D, therefore, Frame D is regenerated and
passed to Station B. Note that the token (T3) is still attached to Frame D since no other
station has had a message to transmit.
FDDI operation, continued
Station A
Station F
Station E
Station D
Station C
Station B
T3 Frame D
Frame A - copy
Frame A - original
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
36
Station B is the destination for Frame D so its makes a copy of Frame D and passes the
frame to Station C. Station C regenerates the frame and passes it to Station D.
FDDI operation, continued
Station A
Station F
Station E
Station D
Station C
Station B
T3 Frame D - original
Frame A - copy
Frame A - original
Frame D - copy
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
37
When Station D receives Frame D, which it originally sent, it removes the frame from the
channel. The unused token (T3) is put back onto the transmission channel and will be
passed from station to station, until one of them requires the token to send a message.
Station A
Station F
Station E
Station D
Station C
Station B
T3
Frame D - original
Frame A - copy
Frame A - original
Frame D - copy
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
38
The FDDI model
Introduction
The FDDI model is made up of three sublayersthe Physical Medium Dependent, Physical
and Media Access Control sublayersand specifications for Station Management. These
layers fit into the Physical layer and Media Access Control sublayer of the OSI model.
Above these sublayers is the IEEE 802.2 Logical Link Control layer. This layer can act as a
bridge and transfer packet between FDDI, Ethernet and Token-ring.
The illustration below shows the relationship between the FDDI model and the OSI model.
FIGURE 10.7:
FDDI AND THE
OSI MODEL
Station
Management
(SMT)
Physical Medium
Dependent (PMD)
Physical Layer Protocol
(PHY)
Media Access Control
(MAC)
Logical Link Control (LLC) - IEEE 802.2
FDDI Model
.............................
.............................
.............................
OSI Physical layer
OSI Data Link layer
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
39
FDDI layer specifications
The PMD specification
The PMD or Physical Medium Dependent specification defines the requirements for the
transmission channel and the devices which connect to the channel. Specifications include
the following:
Optical fiber.
The original specification was for multimode optical fiber cabling with an emphasis
on using 62.5/125 m optical fiber. Later enhancements provided support for the
use of single-mode optical fiber and twisted-pair copper cabling.
Optical fiber connectors.
FDDI optical fiber connectors are referred to as MICs or Media Interface
Connectors. These are two-fiber connectors keyed to prevent a reversal of the
transmit and receive fibers when attaching to devices.
Optical bypass switch.
The optical bypass switch is an optional component. It is used to prevent ring
recovery operations in the event of a device power failure.
Optical transmitters and receivers.
The PMD specification defines the output power of the transmitter, the sensitivity of
the receiver, as well as other details such as output waveforms, center wavelength
and jitter. A maximum of 11 dB loss is allowed on any link of multimode optical
fiber cable.
FDDI layer specifications,
continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
40
The PHY specification
The PHY or Physical specification defines the signaling that takes place over the
transmission channel and how data and control symbols are coded. Included in these
specifications are the following:
Encoding and decoding mechanisms.
Clock synchronization mechanisms.
FDDI layer specifications,
continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
41
The MAC specification
The MAC or Media Access Control specification defines the FDDI frame, access
mechanisms to the channel and the flow control for ring traffic. Included in these
specifications are the following:
Frame format.
The MAC specification defines the structure of both the FDDI token and the FDDI
frame.
Media access.
FDDI uses a token-passing protocol similar to the IEEE 802.5 Token-ring media
access protocol.
Timed Token Rotation (TTR) protocol.
The TTR protocol determine the access time for transmitting information on the
channel.
Access methods.
FDDI specifications provide for both synchronous and asynchronous access.
Ring monitoring.
All devices on an FDDI ring actively engage in the monitoring of ring operations.
Each device is able to detect a ring fault and initiate a bypass operation.
FDDI layer specifications,
continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
42
The SMT specification
The SMT or Station Management specification defines the network management features
of FDDI. It allows a device to detect a problem on the ring and to implement the ring
recovery operations. SMT controls and manages the PMD, PHY and MAC activities, and
is present in every FDDI device.
The functions of Station Management include the following:
Connection Management Function (CMF)
Ring Management Function (RMF)
Physical Connection Management Function (PCM)
Entity Coordination Management Function (ECM)
Configuration Management Function (CFT)
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
43
The FDDI token format
The FDDI token is used by a station to place a message onto the transmission channel and
subsequently, to transfer control to another station.
FIGURE 10.8: THE FDDI TOKEN
Ending
Deli miter (ED)
Frame Control
(FC)
Starting
Deli miter (SD)
Preamble
(PA)
The components of the FDDI token are defined as
follows:
The Preamble (PA).
The preamble is a 64-bit sequence used to
synchronize the frame with the stations clock.
The Starting Delimiter (SD).
The Starting Delimiter is an eight-bit pattern
used to indicate the beginning of a frame.
The Frame Control (FC).
The Frame Control is an eight-bit pattern
defining the type of frame and associated
control functions.
The Ending Delimiter (ED).
The Ending Delimiter is a unique eight-bit
pattern indicating the end of a frame.
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
44
The FDDI frame format
The FDDI frame is used to transmit data
through the FDDI network.
FIGURE 10.9: THE FDDI FRAME FORMAT
The components of the FDDI frame are
defined as follows:
The Preamble (PA).
The preamble is a 64-bit sequence used
to synchronize the frame with the
stations clock.
The Starting Delimiter (SD).
The Starting Delimiter is an eight-bit
pattern used to indicate the beginning of
a frame. Together, the preamble and the
starting delimiter make up the Start of
Frame Sequence (SFS).
The Frame Control (FC).
The Frame Control is an eight-bit
pattern defining the type of frame and
associated control functions.
Frame Status (FS)
Ending Delimiter (ED)
Frame Check Sequence (FCS)
Data
Source Address (SA)
Destination Address (DA)
Frame Control (FC)
Starting Delimiter (SD)
Preamble (PA)
FDDI frame format, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
45
The Destination Address (DA).
The Destination Address is the address of the station to which the data is being
sent. This can be a single station address, a group address (multicast) or a
broadcast address. Both 16-bit locally administered and 48-bit universally
administered addresses can be used.
The Source Address (SA).
The Source Address is the address of the station transmitting the message.
The Data.
The Data field can hold either control information or user data. The length of this
field is variable, with a maximum of 4500 bytes.
The Frame Check Sequence (FCS).
The Frame Check Sequence contains a value used to check the frame control,
destination address, source address and data fields for bit errors.
The Ending Delimiter (ED).
The Ending Delimiter is a unique eight-bit pattern indicating the end of a frame.
The Frame Status (FS).
The Frame Status field consists of three four-bit symbols used to indicate that an
error has been detected, the address has been recognized and the frame has been
copied. Together, the ending delimiter and the frame status field make up the End
of Frame Sequence (EFS).
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
46
FDDI backbones
Introduction
Due to its speed and extended reach capabilities, FDDI is an ideal backbone network. In
addition to its high capacity, FDDI can also accommodate both time-sensitive synchronous
transmissions as well as the more common asynchronous frame traffic.
Backbone design issues
To create an FDDI backbone in an existing LAN environment, FDDI concentrators, bridges,
routers or direct ring attachments can be used. The process is usually as follows:
Individual Ethernet and Token-ring LANs are linked to an FDDI network using
bridges or routers.
Servers are equipped with dual-attachment NICs.
Specific stations requiring high-speed access to servers can be equipped with
single-attachment NICs and linked to an FDDI concentrator. The concentrator is
then connected to the FDDI ring.
Backbone design issues, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
47
FIGURE 10.10:
FDDI BACKBONE
Bridge/Router
Ethernet LAN
Token-ring LAN
Bridge/Router
FDDI Concentrator
Server 1
Server 2
FDDI
Dual
Ring
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
48
Other optical fiber network technologies
FDDI II
Overview
The FDDI II standard is designed to be used with applications that are time sensitive, such
as real-time video. FDDI II uses multiplexing techniques to divide the available bandwidth
into multiple dedicated circuits or channels. These circuits are able to transmit either
asynchronous or isochronous traffic.
A total of 16 channels can be created. Bandwidth allocation to a channel is variable
depending on the needs of the station using the channel. The bandwidth allocated to a
channel will be in the range of 6.144 Mbps to 99.072 Mbps.
Channels can be further subdivided to produce a total of 96 individual circuits operating at
63 Kbps.
FDDI and FDDI II are not compatible technologies. For a network to operate as an FDDI II
network, all attached devices must operate as FDDI II devices. If even one station
operates as an FDDI device, the entire network reverts back to an FDDI network.
Therefore, FDDI and FDDI II devices should each be attached to their own network.
FDDI II, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
49
Functionality
An FDDI II network is designed to operate as follows:
Channels are created by dividing the ring into regular, timed slots. The data to be
transmitted is placed into these slots.
Stations are assigned a priority with respect to its need to transmit. A station
transmitting time-sensitive data is given a higher priority than a station sending an
electronic mail message.
The highest priority stations use the number of slots they require to transmit their
data.
Slots that are unused are immediately reallocated to other stations waiting to
transmit. The higher the priority of a station, the sooner it is able to make its
transmission.
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
50
Fibre Channel
Overview
Please note that this technology uses the British spelling for fiber.
Fibre Channel is not a network in the traditional sense. It is a high-speed transmission
channel used to provide interconnectivity for
computing devices in a limited campus-wide
environment. The primary function of Fibre
Channel is to
interconnect peripheral
devices, mass storage
systems, imaging
systems, archiving
systems, mainframe
computers, high-
performance workstations
and other high-speed devices.
FIGURE 10.11:
FIBRE CHANNEL
CONFIGURATION
High-performance workstation
Server
Mass storage device
(Disk array)
Mainframe or minicomputer
Printer
Ethernet LAN
Fibre
Channel
Switch
Token-ring LAN
FDDI LAN
Fibre Channel, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
51
The most distinguishing feature of Fibre Channel technology is its speed. It is able to
provide a bandwidth ranging from 133 Mbps to 1 Gbps.
Fibre Channel is being promoted by the Fibre Channel System Initiative (FCSI), made up
of major vendors. Its main objective is to develop a high-speed, open connection standard
for workstation and peripheral devices. Currently, the ANSI X3T9 committee is developing
the Fibre Channel Interconnect standard. Once ratified, it will also be added as an ISO
(International Organization for Standardization) standard.
Functionality
A Fibre Channel interface dedicates circuits for the transfer of data. Other devices are
allowed to access the channel when it is free. Simultaneous, multiple sessions are also
accommodated.
Three connection types are possible with Fibre Channel. These are as follows:
Point-to-point connections.
Direct connections between devices through the Fibre Channel switch allows for
high transfer rates over extended distance. Point-to-point connections can take
place over a shared network, however, the cable is unavailable to other users until
the point-to-point communication session is over.
Cluster connections.
Cluster connections are also known as workgroup connections. These are used to
provide communications between high-speed workstations.
Fibre Channel, continued
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
52
Switched LAN connections.
These connections are meant to support Ethernet, Token-ring and FDDI networks.
They permit multiple, simultaneous point-to-point connections between devices.
The Fibre Channel interface supports variable length transmissions. It is able to transmit
data in large blocks without needing to divide it into smaller packets.
Fibre Channel also uses encapsulation techniques to support the interconnection of Fibre
Channel with various network technologies. These techniques add an address header to
the data frames which are then transferred through the Fibre Channel switch. The
technologies which are able to make use of Fibre Channel encapsulation include the
following:
Asynchronous Transfer Mode (ATM).
Fiber Distributed Data Interface (FDDI).
LANs such as Ethernet and Token-ring.
High Performance Parallel Interface (HIPPI).
Intelligent Peripheral Interface (IPI).
Small Computer System Interface (SCSI).
With Fibre Channel, the underlying data formats, packet structures or frame types are not
important to the switching scheme.
Chapter 10 - FDDI Backbone Design
1996, BICSI LAN Design Manual - CD-ROM, Issue 1
53
Optical fiber backbones ........................................................ 1
Overview............................................................................................ 1
Selection criteria .............................................................................. 2
Selecting the optical fiber .................................................................. 3
Number of fibers ............................................................................. 3
Backbone cabling distance.............................................................. 4
Performance criteria........................................................................ 5
Operating wavelength ......................................................................................... 5
Attenuation .......................................................................................................... 6
Bandwidth ........................................................................................................... 8
Optical-electronic equipment ............................................................ 9
Light sources ...................................................................................................... 9
Receivers .......................................................................................................... 10
Link loss budgets .......................................................................... 11
Overview .......................................................................................... 11
Fiber Distributed Data Interface (FDDI) overview ............ 15
FDDI defined ................................................................................... 15
FDDI networking................................................................... 17
Introduction..................................................................................... 17
Applications.................................................................................... 18
FDDI characteristics...................................................................... 20
FDDI terminology........................................................................... 22
FDDI functionality .......................................................................... 25
FDDI operation ............................................................................... 29
C
H
A
P
T
E
R

1
0

-

F
D
D
I

B
A
C
K
B
O
N
E

D
E
S
I
G
N

T
A
B
L
E

O
F

C
O
N
T
E
N
T
S

T
A
B
L
E

O
F

C
O
N
T
E
N
T
S

Chapter 10 - FDDI Backbone Design

1996, BICSI LAN Design Manual - CD-ROM, Issue 1
54
The FDDI model .................................................................... 38
Introduction..................................................................................... 38
FDDI layer specifications ............................................................. 39
The PMD specification .................................................................... 39
The PHY specification .................................................................... 40
The MAC specification .................................................................... 41
The SMT specification .................................................................... 42
The FDDI token format .................................................................. 43
The FDDI frame format ................................................................. 44
FDDI backbones ................................................................... 46
Introduction..................................................................................... 46
Backbone design issues .............................................................. 46
Other optical fiber network technologies ......................... 48
FDDI II .............................................................................................. 48
Overview ......................................................................................... 48
Functionality .................................................................................... 49
Fibre Channel ................................................................................. 50
Overview ......................................................................................... 50
Functionality .................................................................................... 51
C
H
A
P
T
E
R

1
0

-

F
D
D
I

B
A
C
K
B
O
N
E

D
E
S
I
G
N

T
A
B
L
E

O
F

C
O
N
T
E
N
T
S

T
A
B
L
E

O
F

C
O
N
T
E
N
T
S

Chapter 10 - FDDI Backbone Design

1996, BICSI LAN Design Manual - CD-ROM, Issue 1
55
Example 10.1: Link Loss Budgeting ........................................ 12
Example 10.2: FDDI transmission ............................................ 29
C
H
A
P
T
E
R

1
0

-

F
D
D
I

B
A
C
K
B
O
N
E

D
E
S
I
G
N

E
X
A
M
P
L
E
S

E
X
A
M
P
L
E
S

E
X
A
M
P
L
E
S

Chapter 10 - FDDI Backbone Design

1996, BICSI LAN Design Manual - CD-ROM, Issue 1
56
Figure 10.1: Link Loss Budgeting .......................................... 14
Figure 10.2: FDDI configuration ............................................. 24
Figure 10.3: Dual ring operation ............................................. 25
Figure 10.4: FDDI recovery from a cable break ................... 26
Figure 10.5: FDDI recovery from a station failure ................. 27
Figure 10.6: Repaired FDDI ring ............................................ 28
Figure 10.7: FDDI and the OSI model ................................... 38
Figure 10.8: The FDDI token .................................................. 43
Figure 10.9: The FDDI frame format ...................................... 44
Figure 10.10: FDDI backbone .................................................. 47
Figure 10.11: Fibre Channel configuration .............................. 50
C
H
A
P
T
E
R

1
0

-

F
D
D
I

B
A
C
K
B
O
N
E

D
E
S
I
G
N

F
I
G
U
R
E
S

F
I
G
U
R
E
S

F
I
G
U
R
E
S

F
I
G
U
R
E
S